diff options
Diffstat (limited to 'encoding.c')
| -rw-r--r-- | encoding.c | 769 |
1 files changed, 522 insertions, 247 deletions
diff --git a/encoding.c b/encoding.c index 8bfab73177..2737108379 100644 --- a/encoding.c +++ b/encoding.c @@ -17,7 +17,6 @@ #include "internal.h" #include "internal/enc.h" #include "internal/encoding.h" -#include "internal/error.h" #include "internal/inits.h" #include "internal/load.h" #include "internal/object.h" @@ -50,14 +49,16 @@ void rb_encdb_declare(const char *name); int rb_encdb_replicate(const char *name, const char *orig); int rb_encdb_dummy(const char *name); int rb_encdb_alias(const char *alias, const char *orig); +void rb_encdb_set_unicode(int index); #pragma GCC visibility pop #endif static ID id_encoding; VALUE rb_cEncoding; -#define ENCODING_LIST_CAPA 256 -static VALUE rb_encoding_list; +#define DEFAULT_ENCODING_LIST_CAPA 128 +static VALUE rb_default_encoding_list; +static VALUE rb_additional_encoding_list; struct rb_encoding_entry { const char *name; @@ -66,8 +67,9 @@ struct rb_encoding_entry { }; static struct enc_table { - struct rb_encoding_entry list[ENCODING_LIST_CAPA]; + struct rb_encoding_entry *list; int count; + int size; st_table *names; } global_enc_table; @@ -102,7 +104,7 @@ static rb_encoding *global_enc_ascii, static const rb_data_type_t encoding_data_type = { "encoding", {0, 0, 0,}, - 0, 0, RUBY_TYPED_FREE_IMMEDIATELY | RUBY_TYPED_WB_PROTECTED + 0, 0, RUBY_TYPED_FREE_IMMEDIATELY }; #define is_data_encoding(obj) (RTYPEDDATA_P(obj) && RTYPEDDATA_TYPE(obj) == &encoding_data_type) @@ -126,25 +128,47 @@ enc_new(rb_encoding *encoding) static void enc_list_update(int index, rb_raw_encoding *encoding) { - RUBY_ASSERT(index < ENCODING_LIST_CAPA); - - VALUE list = rb_encoding_list; - if (list && NIL_P(rb_ary_entry(list, index))) { - /* initialize encoding data */ - rb_ary_store(list, index, enc_new(encoding)); + if (index < DEFAULT_ENCODING_LIST_CAPA) { + VALUE list = rb_default_encoding_list; + if (list && NIL_P(rb_ary_entry(list, index))) { + /* initialize encoding data */ + rb_ary_store(list, index, enc_new(encoding)); + } + } + else { + RB_VM_LOCK_ENTER(); + { + VALUE list = rb_additional_encoding_list; + if (list && NIL_P(rb_ary_entry(list, index))) { + /* initialize encoding data */ + rb_ary_store(list, index - DEFAULT_ENCODING_LIST_CAPA, enc_new(encoding)); + } + } + RB_VM_LOCK_LEAVE(); } } static VALUE enc_list_lookup(int idx) { - VALUE list, enc = Qnil; + VALUE list, enc; - if (idx < ENCODING_LIST_CAPA) { - list = rb_encoding_list; - RUBY_ASSERT(list); + if (idx < DEFAULT_ENCODING_LIST_CAPA) { + if (!(list = rb_default_encoding_list)) { + rb_bug("rb_enc_from_encoding_index(%d): no rb_default_encoding_list", idx); + } enc = rb_ary_entry(list, idx); } + else { + RB_VM_LOCK_ENTER(); + { + if (!(list = rb_additional_encoding_list)) { + rb_bug("rb_enc_from_encoding_index(%d): no rb_additional_encoding_list", idx); + } + enc = rb_ary_entry(list, idx - DEFAULT_ENCODING_LIST_CAPA); + } + RB_VM_LOCK_LEAVE(); + } if (NIL_P(enc)) { rb_bug("rb_enc_from_encoding_index(%d): not created yet", idx); @@ -186,7 +210,7 @@ check_encoding(rb_encoding *enc) { int index = rb_enc_to_index(enc); if (rb_enc_from_index(index) != enc) - return -1; + return -1; if (rb_enc_autoload_p(enc)) { index = rb_enc_autoload(enc); } @@ -197,7 +221,7 @@ static int enc_check_encoding(VALUE obj) { if (!is_obj_encoding(obj)) { - return -1; + return -1; } return check_encoding(RDATA(obj)->data); } @@ -207,7 +231,7 @@ static void not_encoding(VALUE enc) { rb_raise(rb_eTypeError, "wrong argument type %"PRIsVALUE" (expected Encoding)", - rb_obj_class(enc)); + rb_obj_class(enc)); } static rb_encoding * @@ -215,7 +239,7 @@ must_encoding(VALUE enc) { int index = enc_check_encoding(enc); if (index < 0) { - not_encoding(enc); + not_encoding(enc); } return DATA_PTR(enc); } @@ -225,16 +249,16 @@ must_encindex(int index) { rb_encoding *enc = rb_enc_from_index(index); if (!enc) { - rb_raise(rb_eEncodingError, "encoding index out of bound: %d", - index); + rb_raise(rb_eEncodingError, "encoding index out of bound: %d", + index); } if (ENC_TO_ENCINDEX(enc) != (int)(index & ENC_INDEX_MASK)) { - rb_raise(rb_eEncodingError, "wrong encoding index %d for %s (expected %d)", - index, rb_enc_name(enc), ENC_TO_ENCINDEX(enc)); + rb_raise(rb_eEncodingError, "wrong encoding index %d for %s (expected %d)", + index, rb_enc_name(enc), ENC_TO_ENCINDEX(enc)); } if (rb_enc_autoload_p(enc) && rb_enc_autoload(enc) == -1) { - rb_loaderror("failed to load encoding (%s)", - rb_enc_name(enc)); + rb_loaderror("failed to load encoding (%s)", + rb_enc_name(enc)); } return enc; } @@ -247,16 +271,16 @@ rb_to_encoding_index(VALUE enc) idx = enc_check_encoding(enc); if (idx >= 0) { - return idx; + return idx; } else if (NIL_P(enc = rb_check_string_type(enc))) { - return -1; + return -1; } if (!rb_enc_asciicompat(rb_enc_get(enc))) { - return -1; + return -1; } if (!(name = rb_str_to_cstr(enc))) { - return -1; + return -1; } return rb_enc_find_index(name); } @@ -268,10 +292,10 @@ name_for_encoding(volatile VALUE *enc) const char *n; if (!rb_enc_asciicompat(rb_enc_get(name))) { - rb_raise(rb_eArgError, "invalid encoding name (non ASCII)"); + rb_raise(rb_eArgError, "invalid encoding name (non ASCII)"); } if (!(n = rb_str_to_cstr(name))) { - rb_raise(rb_eArgError, "invalid encoding name (NUL byte)"); + rb_raise(rb_eArgError, "invalid encoding name (NUL byte)"); } return n; } @@ -290,7 +314,7 @@ str_to_encindex(VALUE enc) { int idx = str_find_encindex(enc); if (idx < 0) { - rb_raise(rb_eArgError, "unknown encoding name - %"PRIsVALUE, enc); + rb_raise(rb_eArgError, "unknown encoding name - %"PRIsVALUE, enc); } return idx; } @@ -321,10 +345,16 @@ rb_find_encoding(VALUE enc) static int enc_table_expand(struct enc_table *enc_table, int newsize) { - if (newsize > ENCODING_LIST_CAPA) { - rb_raise(rb_eEncodingError, "too many encoding (> %d)", ENCODING_LIST_CAPA); - } - return newsize; + struct rb_encoding_entry *ent; + int count = newsize; + + if (enc_table->size >= newsize) return newsize; + newsize = (newsize + 7) / 8 * 8; + ent = REALLOC_N(enc_table->list, struct rb_encoding_entry, newsize); + memset(ent + enc_table->size, 0, sizeof(*ent)*(newsize - enc_table->size)); + enc_table->list = ent; + enc_table->size = newsize; + return count; } static int @@ -335,20 +365,20 @@ enc_register_at(struct enc_table *enc_table, int index, const char *name, rb_enc if (!valid_encoding_name_p(name)) return -1; if (!ent->name) { - ent->name = name = strdup(name); + ent->name = name = strdup(name); } else if (STRCASECMP(name, ent->name)) { - return -1; + return -1; } encoding = (rb_raw_encoding *)ent->enc; if (!encoding) { - encoding = xmalloc(sizeof(rb_encoding)); + encoding = xmalloc(sizeof(rb_encoding)); } if (base_encoding) { - *encoding = *base_encoding; + *encoding = *base_encoding; } else { - memset(encoding, 0, sizeof(*ent->enc)); + memset(encoding, 0, sizeof(*ent->enc)); } encoding->name = name; encoding->ruby_encoding_index = index; @@ -375,7 +405,7 @@ static rb_encoding * enc_from_index(struct enc_table *enc_table, int index) { if (UNLIKELY(index < 0 || enc_table->count <= (index &= ENC_INDEX_MASK))) { - return 0; + return 0; } return enc_table->list[index].enc; } @@ -383,7 +413,17 @@ enc_from_index(struct enc_table *enc_table, int index) rb_encoding * rb_enc_from_index(int index) { - return enc_from_index(&global_enc_table, index); + rb_encoding *enc; + + switch (index) { + case ENCINDEX_ASCII: return global_enc_ascii; + case ENCINDEX_UTF_8: return global_enc_utf_8; + case ENCINDEX_US_ASCII: return global_enc_us_ascii; + default: + GLOBAL_ENC_TABLE_EVAL(enc_table, + enc = enc_from_index(enc_table, index)); + return enc; + } } int @@ -422,9 +462,9 @@ enc_registered(struct enc_table *enc_table, const char *name) st_data_t idx = 0; if (!name) return -1; - if (!enc_table->names) return -1; + if (!enc_table->list) return -1; if (st_lookup(enc_table->names, (st_data_t)name, &idx)) { - return (int)idx; + return (int)idx; } return -1; } @@ -444,13 +484,10 @@ rb_encdb_declare(const char *name) } static void -enc_check_addable(struct enc_table *enc_table, const char *name) +enc_check_duplication(struct enc_table *enc_table, const char *name) { if (enc_registered(enc_table, name) >= 0) { - rb_raise(rb_eArgError, "encoding %s is already registered", name); - } - else if (!valid_encoding_name_p(name)) { - rb_raise(rb_eArgError, "invalid encoding name: %s", name); + rb_raise(rb_eArgError, "encoding %s is already registered", name); } } @@ -487,7 +524,11 @@ rb_enc_set_base(const char *name, const char *orig) int rb_enc_set_dummy(int index) { - rb_encoding *enc = global_enc_table.list[index].enc; + rb_encoding *enc; + + GLOBAL_ENC_TABLE_EVAL(enc_table, + enc = enc_table->list[index].enc); + ENC_SET_DUMMY((rb_raw_encoding *)enc); return index; } @@ -497,7 +538,7 @@ enc_replicate(struct enc_table *enc_table, const char *name, rb_encoding *encodi { int idx; - enc_check_addable(enc_table, name); + enc_check_duplication(enc_table, name); idx = enc_register(enc_table, name, encoding); if (idx < 0) rb_raise(rb_eArgError, "invalid encoding name: %s", name); set_base_encoding(enc_table, idx, encoding); @@ -505,6 +546,34 @@ enc_replicate(struct enc_table *enc_table, const char *name, rb_encoding *encodi return idx; } +int +rb_enc_replicate(const char *name, rb_encoding *encoding) +{ + int r; + + GLOBAL_ENC_TABLE_EVAL(enc_table, + r = enc_replicate(enc_table, name, encoding)); + + return r; +} + +/* + * call-seq: + * enc.replicate(name) -> encoding + * + * Returns a replicated encoding of _enc_ whose name is _name_. + * The new encoding should have the same byte structure of _enc_. + * If _name_ is used by another encoding, raise ArgumentError. + * + */ +static VALUE +enc_replicate_m(VALUE encoding, VALUE name) +{ + int idx = rb_enc_replicate(name_for_encoding(&name), rb_to_encoding(encoding)); + RB_GC_GUARD(name); + return rb_enc_from_encoding_index(idx); +} + static int enc_replicate_with_index(struct enc_table *enc_table, const char *name, rb_encoding *origenc, int idx) { @@ -512,11 +581,11 @@ enc_replicate_with_index(struct enc_table *enc_table, const char *name, rb_encod idx = enc_register(enc_table, name, origenc); } else { - idx = enc_register_at(enc_table, idx, name, origenc); + idx = enc_register_at(enc_table, idx, name, origenc); } if (idx >= 0) { - set_base_encoding(enc_table, idx, origenc); - set_encoding_const(name, rb_enc_from_index(idx)); + set_base_encoding(enc_table, idx, origenc); + set_encoding_const(name, rb_enc_from_index(idx)); } else { rb_raise(rb_eArgError, "failed to replicate encoding"); @@ -636,7 +705,7 @@ static int enc_alias_internal(struct enc_table *enc_table, const char *alias, int idx) { return st_insert2(enc_table->names, (st_data_t)alias, (st_data_t)idx, - enc_dup_name); + enc_dup_name); } static int @@ -644,7 +713,7 @@ enc_alias(struct enc_table *enc_table, const char *alias, int idx) { if (!valid_encoding_name_p(alias)) return -1; if (!enc_alias_internal(enc_table, alias, idx)) - set_encoding_const(alias, enc_from_index(enc_table, idx)); + set_encoding_const(alias, enc_from_index(enc_table, idx)); return idx; } @@ -655,7 +724,7 @@ rb_enc_alias(const char *alias, const char *orig) GLOBAL_ENC_TABLE_ENTER(enc_table); { - enc_check_addable(enc_table, alias); + enc_check_duplication(enc_table, alias); if ((idx = rb_enc_find_index(orig)) < 0) { r = -1; } @@ -687,23 +756,29 @@ rb_encdb_alias(const char *alias, const char *orig) return r; } +void +rb_encdb_set_unicode(int index) +{ + rb_raw_encoding *enc = (rb_raw_encoding *)rb_enc_from_index(index); + ASSUME(enc); + enc->flags |= ONIGENC_FLAG_UNICODE; +} + static void rb_enc_init(struct enc_table *enc_table) { enc_table_expand(enc_table, ENCODING_COUNT + 1); if (!enc_table->names) { - enc_table->names = st_init_strcasetable_with_size(ENCODING_LIST_CAPA); + enc_table->names = st_init_strcasetable(); } -#define OnigEncodingASCII_8BIT OnigEncodingASCII #define ENC_REGISTER(enc) enc_register_at(enc_table, ENCINDEX_##enc, rb_enc_name(&OnigEncoding##enc), &OnigEncoding##enc) - ENC_REGISTER(ASCII_8BIT); + ENC_REGISTER(ASCII); ENC_REGISTER(UTF_8); ENC_REGISTER(US_ASCII); - global_enc_ascii = enc_table->list[ENCINDEX_ASCII_8BIT].enc; + global_enc_ascii = enc_table->list[ENCINDEX_ASCII].enc; global_enc_utf_8 = enc_table->list[ENCINDEX_UTF_8].enc; global_enc_us_ascii = enc_table->list[ENCINDEX_US_ASCII].enc; #undef ENC_REGISTER -#undef OnigEncodingASCII_8BIT #define ENCDB_REGISTER(name, enc) enc_register_at(enc_table, ENCINDEX_##enc, name, NULL) ENCDB_REGISTER("UTF-16BE", UTF_16BE); ENCDB_REGISTER("UTF-16LE", UTF_16LE); @@ -738,9 +813,9 @@ load_encoding(const char *name) int idx; while (s < e) { - if (!ISALNUM(*s)) *s = '_'; - else if (ISUPPER(*s)) *s = (char)TOLOWER(*s); - ++s; + if (!ISALNUM(*s)) *s = '_'; + else if (ISUPPER(*s)) *s = (char)TOLOWER(*s); + ++s; } enclib = rb_fstring(enclib); ruby_debug = Qfalse; @@ -773,16 +848,16 @@ enc_autoload_body(struct enc_table *enc_table, rb_encoding *enc) if (base) { int i = 0; - do { - if (i >= enc_table->count) return -1; - } while (enc_table->list[i].enc != base && (++i, 1)); - if (rb_enc_autoload_p(base)) { - if (rb_enc_autoload(base) < 0) return -1; - } - i = enc->ruby_encoding_index; - enc_register_at(enc_table, i & ENC_INDEX_MASK, rb_enc_name(enc), base); + do { + if (i >= enc_table->count) return -1; + } while (enc_table->list[i].enc != base && (++i, 1)); + if (rb_enc_autoload_p(base)) { + if (rb_enc_autoload(base) < 0) return -1; + } + i = enc->ruby_encoding_index; + enc_register_at(enc_table, i & ENC_INDEX_MASK, rb_enc_name(enc), base); ((rb_raw_encoding *)enc)->ruby_encoding_index = i; - i &= ENC_INDEX_MASK; + i &= ENC_INDEX_MASK; return i; } else { @@ -805,23 +880,25 @@ rb_enc_autoload(rb_encoding *enc) int rb_enc_find_index(const char *name) { - int i = enc_registered(&global_enc_table, name); + int i; rb_encoding *enc; + GLOBAL_ENC_TABLE_EVAL(enc_table, i = enc_registered(enc_table, name)); + if (i < 0) { - i = load_encoding(name); + i = load_encoding(name); } else if (!(enc = rb_enc_from_index(i))) { - if (i != UNSPECIFIED_ENCODING) { - rb_raise(rb_eArgError, "encoding %s is not registered", name); - } + if (i != UNSPECIFIED_ENCODING) { + rb_raise(rb_eArgError, "encoding %s is not registered", name); + } } else if (rb_enc_autoload_p(enc)) { - if (rb_enc_autoload(enc) < 0) { - rb_warn("failed to load encoding (%s); use ASCII-8BIT instead", - name); - return 0; - } + if (rb_enc_autoload(enc) < 0) { + rb_warn("failed to load encoding (%s); use ASCII-8BIT instead", + name); + return 0; + } } return i; } @@ -854,11 +931,11 @@ enc_capable(VALUE obj) case T_REGEXP: case T_FILE: case T_SYMBOL: - return TRUE; + return TRUE; case T_DATA: - if (is_data_encoding(obj)) return TRUE; + if (is_data_encoding(obj)) return TRUE; default: - return FALSE; + return FALSE; } } @@ -880,11 +957,11 @@ enc_get_index_str(VALUE str) { int i = ENCODING_GET_INLINED(str); if (i == ENCODING_INLINE_MAX) { - VALUE iv; + VALUE iv; #if 0 - iv = rb_ivar_get(str, rb_id_encoding()); - i = NUM2INT(iv); + iv = rb_ivar_get(str, rb_id_encoding()); + i = NUM2INT(iv); #else /* * Tentatively, assume ASCII-8BIT, if encoding index instance @@ -892,7 +969,7 @@ enc_get_index_str(VALUE str) * all instance variables are removed in `obj_free`. */ iv = rb_attr_get(str, rb_id_encoding()); - i = NIL_P(iv) ? ENCINDEX_ASCII_8BIT : NUM2INT(iv); + i = NIL_P(iv) ? ENCINDEX_ASCII : NUM2INT(iv); #endif } return i; @@ -905,31 +982,31 @@ rb_enc_get_index(VALUE obj) VALUE tmp; if (SPECIAL_CONST_P(obj)) { - if (!SYMBOL_P(obj)) return -1; - obj = rb_sym2str(obj); + if (!SYMBOL_P(obj)) return -1; + obj = rb_sym2str(obj); } switch (BUILTIN_TYPE(obj)) { case T_STRING: case T_SYMBOL: case T_REGEXP: - i = enc_get_index_str(obj); - break; + i = enc_get_index_str(obj); + break; case T_FILE: - tmp = rb_funcallv(obj, rb_intern("internal_encoding"), 0, 0); - if (NIL_P(tmp)) { - tmp = rb_funcallv(obj, rb_intern("external_encoding"), 0, 0); - } - if (is_obj_encoding(tmp)) { - i = enc_check_encoding(tmp); - } - break; + tmp = rb_funcallv(obj, rb_intern("internal_encoding"), 0, 0); + if (NIL_P(tmp)) { + tmp = rb_funcallv(obj, rb_intern("external_encoding"), 0, 0); + } + if (is_obj_encoding(tmp)) { + i = enc_check_encoding(tmp); + } + break; case T_DATA: - if (is_data_encoding(obj)) { - i = enc_check_encoding(obj); - } - break; + if (is_data_encoding(obj)) { + i = enc_check_encoding(obj); + } + break; default: - break; + break; } return i; } @@ -942,8 +1019,8 @@ enc_set_index(VALUE obj, int idx) } if (idx < ENCODING_INLINE_MAX) { - ENCODING_SET_INLINED(obj, idx); - return; + ENCODING_SET_INLINED(obj, idx); + return; } ENCODING_SET_INLINED(obj, ENCODING_INLINE_MAX); rb_ivar_set(obj, rb_id_encoding(), INT2NUM(idx)); @@ -967,19 +1044,19 @@ rb_enc_associate_index(VALUE obj, int idx) rb_check_frozen(obj); oldidx = rb_enc_get_index(obj); if (oldidx == idx) - return obj; + return obj; if (SPECIAL_CONST_P(obj)) { - rb_raise(rb_eArgError, "cannot set encoding"); + rb_raise(rb_eArgError, "cannot set encoding"); } enc = must_encindex(idx); if (!ENC_CODERANGE_ASCIIONLY(obj) || - !rb_enc_asciicompat(enc)) { - ENC_CODERANGE_CLEAR(obj); + !rb_enc_asciicompat(enc)) { + ENC_CODERANGE_CLEAR(obj); } termlen = rb_enc_mbminlen(enc); oldtermlen = rb_enc_mbminlen(rb_enc_from_index(oldidx)); if (oldtermlen != termlen && RB_TYPE_P(obj, T_STRING)) { - rb_str_change_terminator_length(obj, oldtermlen, termlen); + rb_str_change_terminator_length(obj, oldtermlen, termlen); } enc_set_index(obj, idx); return obj; @@ -1001,9 +1078,9 @@ static rb_encoding* rb_encoding_check(rb_encoding* enc, VALUE str1, VALUE str2) { if (!enc) - rb_raise(rb_eEncCompatError, "incompatible character encodings: %s and %s", - rb_enc_name(rb_enc_get(str1)), - rb_enc_name(rb_enc_get(str2))); + rb_raise(rb_eEncCompatError, "incompatible character encodings: %s and %s", + rb_enc_name(rb_enc_get(str1)), + rb_enc_name(rb_enc_get(str2))); return enc; } @@ -1032,48 +1109,48 @@ enc_compatible_latter(VALUE str1, VALUE str2, int idx1, int idx2) isstr2 = RB_TYPE_P(str2, T_STRING); if (isstr2 && RSTRING_LEN(str2) == 0) - return enc1; + return enc1; isstr1 = RB_TYPE_P(str1, T_STRING); if (isstr1 && isstr2 && RSTRING_LEN(str1) == 0) - return (rb_enc_asciicompat(enc1) && rb_enc_str_asciionly_p(str2)) ? enc1 : enc2; + return (rb_enc_asciicompat(enc1) && rb_enc_str_asciionly_p(str2)) ? enc1 : enc2; if (!rb_enc_asciicompat(enc1) || !rb_enc_asciicompat(enc2)) { - return 0; + return 0; } /* objects whose encoding is the same of contents */ if (!isstr2 && idx2 == ENCINDEX_US_ASCII) - return enc1; + return enc1; if (!isstr1 && idx1 == ENCINDEX_US_ASCII) - return enc2; + return enc2; if (!isstr1) { - VALUE tmp = str1; - int idx0 = idx1; - str1 = str2; - str2 = tmp; - idx1 = idx2; - idx2 = idx0; - idx0 = isstr1; - isstr1 = isstr2; - isstr2 = idx0; + VALUE tmp = str1; + int idx0 = idx1; + str1 = str2; + str2 = tmp; + idx1 = idx2; + idx2 = idx0; + idx0 = isstr1; + isstr1 = isstr2; + isstr2 = idx0; } if (isstr1) { - int cr1, cr2; - - cr1 = rb_enc_str_coderange(str1); - if (isstr2) { - cr2 = rb_enc_str_coderange(str2); - if (cr1 != cr2) { - /* may need to handle ENC_CODERANGE_BROKEN */ - if (cr1 == ENC_CODERANGE_7BIT) return enc2; - if (cr2 == ENC_CODERANGE_7BIT) return enc1; - } - if (cr2 == ENC_CODERANGE_7BIT) { - return enc1; - } - } - if (cr1 == ENC_CODERANGE_7BIT) - return enc2; + int cr1, cr2; + + cr1 = rb_enc_str_coderange(str1); + if (isstr2) { + cr2 = rb_enc_str_coderange(str2); + if (cr1 != cr2) { + /* may need to handle ENC_CODERANGE_BROKEN */ + if (cr1 == ENC_CODERANGE_7BIT) return enc2; + if (cr2 == ENC_CODERANGE_7BIT) return enc1; + } + if (cr2 == ENC_CODERANGE_7BIT) { + return enc1; + } + } + if (cr1 == ENC_CODERANGE_7BIT) + return enc2; } return 0; } @@ -1088,10 +1165,10 @@ enc_compatible_str(VALUE str1, VALUE str2) return 0; if (idx1 == idx2) { - return rb_enc_from_index(idx1); + return rb_enc_from_index(idx1); } else { - return enc_compatible_latter(str1, str2, idx1, idx2); + return enc_compatible_latter(str1, str2, idx1, idx2); } } @@ -1105,7 +1182,7 @@ rb_enc_compatible(VALUE str1, VALUE str2) return 0; if (idx1 == idx2) { - return rb_enc_from_index(idx1); + return rb_enc_from_index(idx1); } return enc_compatible_latter(str1, str2, idx1, idx2); @@ -1130,7 +1207,7 @@ rb_obj_encoding(VALUE obj) { int idx = rb_enc_get_index(obj); if (idx < 0) { - rb_raise(rb_eTypeError, "unknown encoding"); + rb_raise(rb_eTypeError, "unknown encoding"); } return rb_enc_from_encoding_index(idx & ENC_INDEX_MASK); } @@ -1197,7 +1274,7 @@ rb_enc_codepoint_len(const char *p, const char *e, int *len_p, rb_encoding *enc) rb_raise(rb_eArgError, "empty string"); r = rb_enc_precise_mbclen(p, e, enc); if (!MBCLEN_CHARFOUND_P(r)) { - rb_raise(rb_eArgError, "invalid byte sequence in %s", rb_enc_name(enc)); + rb_raise(rb_eArgError, "invalid byte sequence in %s", rb_enc_name(enc)); } if (len_p) *len_p = MBCLEN_CHARFOUND_LEN(r); return rb_enc_mbc_to_codepoint(p, e, enc); @@ -1208,7 +1285,7 @@ rb_enc_codelen(int c, rb_encoding *enc) { int n = ONIGENC_CODE_TO_MBCLEN(enc,c); if (n == 0) { - rb_raise(rb_eArgError, "invalid codepoint 0x%x in %s", c, rb_enc_name(enc)); + rb_raise(rb_eArgError, "invalid codepoint 0x%x in %s", c, rb_enc_name(enc)); } return n; } @@ -1240,16 +1317,16 @@ enc_inspect(VALUE self) rb_encoding *enc; if (!is_data_encoding(self)) { - not_encoding(self); + not_encoding(self); } if (!(enc = DATA_PTR(self)) || rb_enc_from_index(rb_enc_to_index(enc)) != enc) { - rb_raise(rb_eTypeError, "broken Encoding"); + rb_raise(rb_eTypeError, "broken Encoding"); } return rb_enc_sprintf(rb_usascii_encoding(), - "#<%"PRIsVALUE":%s%s%s>", rb_obj_class(self), - rb_enc_name(enc), - (ENC_DUMMY_P(enc) ? " (dummy)" : ""), - rb_enc_autoload_p(enc) ? " (autoload)" : ""); + "#<%"PRIsVALUE":%s%s%s>", rb_obj_class(self), + rb_enc_name(enc), + (ENC_DUMMY_P(enc) ? " (dummy)" : ""), + rb_enc_autoload_p(enc) ? " (autoload)" : ""); } /* @@ -1273,8 +1350,8 @@ enc_names_i(st_data_t name, st_data_t idx, st_data_t args) VALUE *arg = (VALUE *)args; if ((int)idx == (int)arg[0]) { - VALUE str = rb_fstring_cstr((char *)name); - rb_ary_push(arg[1], str); + VALUE str = rb_fstring_cstr((char *)name); + rb_ary_push(arg[1], str); } return ST_CONTINUE; } @@ -1294,7 +1371,10 @@ enc_names(VALUE self) args[0] = (VALUE)rb_to_encoding_index(self); args[1] = rb_ary_new2(0); - st_foreach(global_enc_table.names, enc_names_i, (st_data_t)args); + + GLOBAL_ENC_TABLE_EVAL(enc_table, + st_foreach(enc_table->names, enc_names_i, (st_data_t)args)); + return args[1]; } @@ -1320,7 +1400,14 @@ static VALUE enc_list(VALUE klass) { VALUE ary = rb_ary_new2(0); - rb_ary_replace(ary, rb_encoding_list); + + RB_VM_LOCK_ENTER(); + { + rb_ary_replace(ary, rb_default_encoding_list); + rb_ary_concat(ary, rb_additional_encoding_list); + } + RB_VM_LOCK_LEAVE(); + return ary; } @@ -1351,7 +1438,7 @@ enc_find(VALUE klass, VALUE enc) { int idx; if (is_obj_encoding(enc)) - return enc; + return enc; idx = str_to_encindex(enc); if (idx == UNSPECIFIED_ENCODING) return Qnil; return rb_enc_from_encoding_index(idx); @@ -1433,7 +1520,7 @@ rb_ascii8bit_encoding(void) int rb_ascii8bit_encindex(void) { - return ENCINDEX_ASCII_8BIT; + return ENCINDEX_ASCII; } rb_encoding * @@ -1469,17 +1556,15 @@ rb_locale_encindex(void) if (idx < 0) idx = ENCINDEX_UTF_8; - if (enc_registered(&global_enc_table, "locale") < 0) { + GLOBAL_ENC_TABLE_ENTER(enc_table); + if (enc_registered(enc_table, "locale") < 0) { # if defined _WIN32 - void Init_w32_codepage(void); - Init_w32_codepage(); + void Init_w32_codepage(void); + Init_w32_codepage(); # endif - GLOBAL_ENC_TABLE_ENTER(enc_table); - { - enc_alias_internal(enc_table, "locale", idx); - } - GLOBAL_ENC_TABLE_LEAVE(); + enc_alias_internal(enc_table, "locale", idx); } + GLOBAL_ENC_TABLE_LEAVE(); return idx; } @@ -1493,8 +1578,13 @@ rb_locale_encoding(void) int rb_filesystem_encindex(void) { - int idx = enc_registered(&global_enc_table, "filesystem"); - if (idx < 0) idx = ENCINDEX_ASCII_8BIT; + int idx; + + GLOBAL_ENC_TABLE_EVAL(enc_table, + idx = enc_registered(enc_table, "filesystem")); + + if (idx < 0) + idx = ENCINDEX_ASCII; return idx; } @@ -1517,8 +1607,8 @@ enc_set_default_encoding(struct default_encoding *def, VALUE encoding, const cha int overridden = FALSE; if (def->index != -2) - /* Already set */ - overridden = TRUE; + /* Already set */ + overridden = TRUE; GLOBAL_ENC_TABLE_ENTER(enc_table); { @@ -1717,45 +1807,45 @@ set_encoding_const(const char *name, rb_encoding *enc) if (ISDIGIT(*s)) return; if (ISUPPER(*s)) { - hasupper = 1; - while (*++s && (ISALNUM(*s) || *s == '_')) { - if (ISLOWER(*s)) haslower = 1; - } + hasupper = 1; + while (*++s && (ISALNUM(*s) || *s == '_')) { + if (ISLOWER(*s)) haslower = 1; + } } if (!*s) { - if (s - name > ENCODING_NAMELEN_MAX) return; - valid = 1; - rb_define_const(rb_cEncoding, name, encoding); + if (s - name > ENCODING_NAMELEN_MAX) return; + valid = 1; + rb_define_const(rb_cEncoding, name, encoding); } if (!valid || haslower) { - size_t len = s - name; - if (len > ENCODING_NAMELEN_MAX) return; - if (!haslower || !hasupper) { - do { - if (ISLOWER(*s)) haslower = 1; - if (ISUPPER(*s)) hasupper = 1; - } while (*++s && (!haslower || !hasupper)); - len = s - name; - } - len += strlen(s); - if (len++ > ENCODING_NAMELEN_MAX) return; - MEMCPY(s = ALLOCA_N(char, len), name, char, len); - name = s; - if (!valid) { - if (ISLOWER(*s)) *s = ONIGENC_ASCII_CODE_TO_UPPER_CASE((int)*s); - for (; *s; ++s) { - if (!ISALNUM(*s)) *s = '_'; - } - if (hasupper) { - rb_define_const(rb_cEncoding, name, encoding); - } - } - if (haslower) { - for (s = (char *)name; *s; ++s) { - if (ISLOWER(*s)) *s = ONIGENC_ASCII_CODE_TO_UPPER_CASE((int)*s); - } - rb_define_const(rb_cEncoding, name, encoding); - } + size_t len = s - name; + if (len > ENCODING_NAMELEN_MAX) return; + if (!haslower || !hasupper) { + do { + if (ISLOWER(*s)) haslower = 1; + if (ISUPPER(*s)) hasupper = 1; + } while (*++s && (!haslower || !hasupper)); + len = s - name; + } + len += strlen(s); + if (len++ > ENCODING_NAMELEN_MAX) return; + MEMCPY(s = ALLOCA_N(char, len), name, char, len); + name = s; + if (!valid) { + if (ISLOWER(*s)) *s = ONIGENC_ASCII_CODE_TO_UPPER_CASE((int)*s); + for (; *s; ++s) { + if (!ISALNUM(*s)) *s = '_'; + } + if (hasupper) { + rb_define_const(rb_cEncoding, name, encoding); + } + } + if (haslower) { + for (s = (char *)name; *s; ++s) { + if (ISLOWER(*s)) *s = ONIGENC_ASCII_CODE_TO_UPPER_CASE((int)*s); + } + rb_define_const(rb_cEncoding, name, encoding); + } } } @@ -1785,8 +1875,15 @@ rb_enc_name_list_i(st_data_t name, st_data_t idx, st_data_t arg) static VALUE rb_enc_name_list(VALUE klass) { - VALUE ary = rb_ary_new2(global_enc_table.names->num_entries); - st_foreach(global_enc_table.names, rb_enc_name_list_i, (st_data_t)ary); + VALUE ary; + + GLOBAL_ENC_TABLE_ENTER(enc_table); + { + ary = rb_ary_new2(enc_table->names->num_entries); + st_foreach(enc_table->names, rb_enc_name_list_i, (st_data_t)ary); + } + GLOBAL_ENC_TABLE_LEAVE(); + return ary; } @@ -1799,14 +1896,14 @@ rb_enc_aliases_enc_i(st_data_t name, st_data_t orig, st_data_t arg) VALUE key, str = rb_ary_entry(ary, idx); if (NIL_P(str)) { - rb_encoding *enc = rb_enc_from_index(idx); + rb_encoding *enc = rb_enc_from_index(idx); - if (!enc) return ST_CONTINUE; - if (STRCASECMP((char*)name, rb_enc_name(enc)) == 0) { - return ST_CONTINUE; - } - str = rb_fstring_cstr(rb_enc_name(enc)); - rb_ary_store(ary, idx, str); + if (!enc) return ST_CONTINUE; + if (STRCASECMP((char*)name, rb_enc_name(enc)) == 0) { + return ST_CONTINUE; + } + str = rb_fstring_cstr(rb_enc_name(enc)); + rb_ary_store(ary, idx, str); } key = rb_fstring_cstr((char *)name); rb_hash_aset(aliases, key, str); @@ -1832,39 +1929,210 @@ rb_enc_aliases(VALUE klass) aliases[0] = rb_hash_new(); aliases[1] = rb_ary_new(); - st_foreach(global_enc_table.names, rb_enc_aliases_enc_i, (st_data_t)aliases); + GLOBAL_ENC_TABLE_EVAL(enc_table, + st_foreach(enc_table->names, rb_enc_aliases_enc_i, (st_data_t)aliases)); return aliases[0]; } /* - * An \Encoding instance represents a character encoding usable in Ruby. - * It is defined as a constant under the \Encoding namespace. - * It has a name and, optionally, aliases: + * An Encoding instance represents a character encoding usable in Ruby. It is + * defined as a constant under the Encoding namespace. It has a name and + * optionally, aliases: + * + * Encoding::ISO_8859_1.name + * #=> "ISO-8859-1" + * + * Encoding::ISO_8859_1.names + * #=> ["ISO-8859-1", "ISO8859-1"] + * + * Ruby methods dealing with encodings return or accept Encoding instances as + * arguments (when a method accepts an Encoding instance as an argument, it + * can be passed an Encoding name or alias instead). + * + * "some string".encoding + * #=> #<Encoding:UTF-8> + * + * string = "some string".encode(Encoding::ISO_8859_1) + * #=> "some string" + * string.encoding + * #=> #<Encoding:ISO-8859-1> + * + * "some string".encode "ISO-8859-1" + * #=> "some string" + * + * Encoding::ASCII_8BIT is a special encoding that is usually used for + * a byte string, not a character string. But as the name insists, its + * characters in the range of ASCII are considered as ASCII + * characters. This is useful when you use ASCII-8BIT characters with + * other ASCII compatible characters. + * + * == Changing an encoding + * + * The associated Encoding of a String can be changed in two different ways. + * + * First, it is possible to set the Encoding of a string to a new Encoding + * without changing the internal byte representation of the string, with + * String#force_encoding. This is how you can tell Ruby the correct encoding + * of a string. + * + * string + * #=> "R\xC3\xA9sum\xC3\xA9" + * string.encoding + * #=> #<Encoding:ISO-8859-1> + * string.force_encoding(Encoding::UTF_8) + * #=> "R\u00E9sum\u00E9" + * + * Second, it is possible to transcode a string, i.e. translate its internal + * byte representation to another encoding. Its associated encoding is also + * set to the other encoding. See String#encode for the various forms of + * transcoding, and the Encoding::Converter class for additional control over + * the transcoding process. + * + * string + * #=> "R\u00E9sum\u00E9" + * string.encoding + * #=> #<Encoding:UTF-8> + * string = string.encode!(Encoding::ISO_8859_1) + * #=> "R\xE9sum\xE9" + * string.encoding + * #=> #<Encoding::ISO-8859-1> + * + * == Script encoding * - * Encoding::US_ASCII.name # => "US-ASCII" - * Encoding::US_ASCII.names # => ["US-ASCII", "ASCII", "ANSI_X3.4-1968", "646"] + * All Ruby script code has an associated Encoding which any String literal + * created in the source code will be associated to. * - * A Ruby method that accepts an encoding as an argument will accept: + * The default script encoding is Encoding::UTF_8 after v2.0, but it + * can be changed by a magic comment on the first line of the source + * code file (or second line, if there is a shebang line on the + * first). The comment must contain the word <code>coding</code> or + * <code>encoding</code>, followed by a colon, space and the Encoding + * name or alias: * - * - An \Encoding object. - * - The name of an encoding. - * - An alias for an encoding name. + * # encoding: UTF-8 * - * These are equivalent: + * "some string".encoding + * #=> #<Encoding:UTF-8> * - * 'foo'.encode(Encoding::US_ASCII) # Encoding object. - * 'foo'.encode('US-ASCII') # Encoding name. - * 'foo'.encode('ASCII') # Encoding alias. + * The <code>__ENCODING__</code> keyword returns the script encoding of the file + * which the keyword is written: * - * For a full discussion of encodings and their uses, - * see {the Encodings document}[rdoc-ref:encodings.rdoc]. + * # encoding: ISO-8859-1 * - * Encoding::ASCII_8BIT is a special-purpose encoding that is usually used for - * a string of bytes, not a string of characters. - * But as the name indicates, its characters in the ASCII range - * are considered as ASCII characters. - * This is useful when you use other ASCII-compatible encodings. + * __ENCODING__ + * #=> #<Encoding:ISO-8859-1> + * + * <code>ruby -K</code> will change the default locale encoding, but this is + * not recommended. Ruby source files should declare its script encoding by a + * magic comment even when they only depend on US-ASCII strings or regular + * expressions. + * + * == Locale encoding + * + * The default encoding of the environment. Usually derived from locale. + * + * see Encoding.locale_charmap, Encoding.find('locale') + * + * == Filesystem encoding + * + * The default encoding of strings from the filesystem of the environment. + * This is used for strings of file names or paths. + * + * see Encoding.find('filesystem') + * + * == External encoding + * + * Each IO object has an external encoding which indicates the encoding that + * Ruby will use to read its data. By default Ruby sets the external encoding + * of an IO object to the default external encoding. The default external + * encoding is set by locale encoding or the interpreter <code>-E</code> option. + * Encoding.default_external returns the current value of the external + * encoding. + * + * ENV["LANG"] + * #=> "UTF-8" + * Encoding.default_external + * #=> #<Encoding:UTF-8> + * + * $ ruby -E ISO-8859-1 -e "p Encoding.default_external" + * #<Encoding:ISO-8859-1> + * + * $ LANG=C ruby -e 'p Encoding.default_external' + * #<Encoding:US-ASCII> + * + * The default external encoding may also be set through + * Encoding.default_external=, but you should not do this as strings created + * before and after the change will have inconsistent encodings. Instead use + * <code>ruby -E</code> to invoke ruby with the correct external encoding. + * + * When you know that the actual encoding of the data of an IO object is not + * the default external encoding, you can reset its external encoding with + * IO#set_encoding or set it at IO object creation (see IO.new options). + * + * == Internal encoding + * + * To process the data of an IO object which has an encoding different + * from its external encoding, you can set its internal encoding. Ruby will use + * this internal encoding to transcode the data when it is read from the IO + * object. + * + * Conversely, when data is written to the IO object it is transcoded from the + * internal encoding to the external encoding of the IO object. + * + * The internal encoding of an IO object can be set with + * IO#set_encoding or at IO object creation (see IO.new options). + * + * The internal encoding is optional and when not set, the Ruby default + * internal encoding is used. If not explicitly set this default internal + * encoding is +nil+ meaning that by default, no transcoding occurs. + * + * The default internal encoding can be set with the interpreter option + * <code>-E</code>. Encoding.default_internal returns the current internal + * encoding. + * + * $ ruby -e 'p Encoding.default_internal' + * nil + * + * $ ruby -E ISO-8859-1:UTF-8 -e "p [Encoding.default_external, \ + * Encoding.default_internal]" + * [#<Encoding:ISO-8859-1>, #<Encoding:UTF-8>] + * + * The default internal encoding may also be set through + * Encoding.default_internal=, but you should not do this as strings created + * before and after the change will have inconsistent encodings. Instead use + * <code>ruby -E</code> to invoke ruby with the correct internal encoding. + * + * == IO encoding example + * + * In the following example a UTF-8 encoded string "R\u00E9sum\u00E9" is transcoded for + * output to ISO-8859-1 encoding, then read back in and transcoded to UTF-8: + * + * string = "R\u00E9sum\u00E9" + * + * open("transcoded.txt", "w:ISO-8859-1") do |io| + * io.write(string) + * end + * + * puts "raw text:" + * p File.binread("transcoded.txt") + * puts + * + * open("transcoded.txt", "r:ISO-8859-1:UTF-8") do |io| + * puts "transcoded text:" + * p io.read + * end + * + * While writing the file, the internal encoding is not specified as it is + * only necessary for reading. While reading the file both the internal and + * external encoding must be specified to obtain the correct result. + * + * $ ruby t.rb + * raw text: + * "R\xE9sum\xE9" + * + * transcoded text: + * "R\u00E9sum\u00E9" * */ @@ -1883,6 +2151,7 @@ Init_Encoding(void) rb_define_method(rb_cEncoding, "names", enc_names, 0); rb_define_method(rb_cEncoding, "dummy?", enc_dummy_p, 0); rb_define_method(rb_cEncoding, "ascii_compatible?", enc_ascii_compatible_p, 0); + rb_define_method(rb_cEncoding, "replicate", enc_replicate_m, 1); rb_define_singleton_method(rb_cEncoding, "list", enc_list, 0); rb_define_singleton_method(rb_cEncoding, "name_list", rb_enc_name_list, 0); rb_define_singleton_method(rb_cEncoding, "aliases", rb_enc_aliases, 0); @@ -1900,12 +2169,18 @@ Init_Encoding(void) struct enc_table *enc_table = &global_enc_table; - list = rb_encoding_list = rb_ary_new2(ENCODING_LIST_CAPA); + if (DEFAULT_ENCODING_LIST_CAPA < enc_table->count) rb_bug("DEFAULT_ENCODING_LIST_CAPA is too small"); + + list = rb_additional_encoding_list = rb_ary_new(); + RBASIC_CLEAR_CLASS(list); + rb_gc_register_mark_object(list); + + list = rb_default_encoding_list = rb_ary_new2(DEFAULT_ENCODING_LIST_CAPA); RBASIC_CLEAR_CLASS(list); rb_gc_register_mark_object(list); for (i = 0; i < enc_table->count; ++i) { - rb_ary_push(list, enc_new(enc_table->list[i].enc)); + rb_ary_push(list, enc_new(enc_table->list[i].enc)); } rb_marshal_define_compat(rb_cEncoding, Qnil, 0, enc_m_loader); @@ -1922,5 +2197,5 @@ Init_encodings(void) void rb_enc_foreach_name(int (*func)(st_data_t name, st_data_t idx, st_data_t arg), st_data_t arg) { - st_foreach(global_enc_table.names, func, arg); + GLOBAL_ENC_TABLE_EVAL(enc_table, st_foreach(enc_table->names, func, arg)); } |