March 2010 - macruby-changes - lists.macosforge.org

[3709] MacRuby/branches/icu/string.c
by source_changes＠macosforge.org 06 Mar '10

06 Mar '10

Revision: 3709 http://trac.macosforge.org/projects/ruby/changeset/3709 Author: lsansonetti(a)apple.com Date: 2010-03-05 16:49:40 -0800 (Fri, 05 Mar 2010) Log Message: ----------- added #partition, #rpartition Modified Paths: -------------- MacRuby/branches/icu/string.c Modified: MacRuby/branches/icu/string.c =================================================================== --- MacRuby/branches/icu/string.c 2010-03-05 23:26:42 UTC (rev 3708) +++ MacRuby/branches/icu/string.c 2010-03-06 00:49:40 UTC (rev 3709) @@ -1637,7 +1637,7 @@ static VALUE rb_str_subpat(VALUE str, VALUE re, int nth) { - if (rb_reg_search(re, str, 0, 0) >= 0) { + if (rb_reg_search(re, str, 0, false) >= 0) { return rb_reg_nth_match(nth, rb_backref_get()); } return Qnil; @@ -4812,6 +4812,95 @@ return LONG2NUM(rb_str_hash(str)); } +/* + * call-seq: + * str.partition(sep) => [head, sep, tail] + * + * Searches the string for sep and returns the part before + * it, the sep, and the part after it. If sep is not found, + * returns str and two empty strings. + * + * "hello".partition("l") #=> ["he", "l", "lo"] + * "hello".partition("x") #=> ["hello", "", ""] + */ + +static VALUE +rstr_partition(VALUE str, SEL sel, VALUE sep) +{ + long pos = 0; + long seplen = 0; + bool regex = false; + + if (TYPE(sep) == T_REGEXP) { + pos = rb_reg_search(sep, str, 0, false); + regex = true; + } + else { + StringValue(sep); + seplen = rb_str_chars_len(sep); + pos = str_index_for_string(RSTR(str), str_need_string(sep), + 0, -1, false, true); + } + if (pos < 0) { +failed: + return rb_ary_new3(3, str, rb_str_new(NULL,0), rb_str_new(NULL,0)); + } + if (regex) { + sep = rb_str_subpat(str, sep, 0); + seplen = rb_str_chars_len(sep); + if (pos == 0 && seplen == 0) { + goto failed; + } + } + const long len = rb_str_chars_len(str); + return rb_ary_new3(3, rstr_substr(str, 0, pos), sep, + rstr_substr(str, pos + seplen, len - pos - seplen)); +} + +/* + * call-seq: + * str.rpartition(sep) => [head, sep, tail] + * + * Searches sep in the string from the end of the string, and + * returns the part before it, the sep, and the part after it. + * If sep is not found, returns two empty strings and + * str. + * + * "hello".rpartition("l") #=> ["hel", "l", "o"] + * "hello".rpartition("x") #=> ["", "", "hello"] + */ + +static VALUE +rstr_rpartition(VALUE str, SEL sel, VALUE sep) +{ + const long len = rb_str_chars_len(str); + long pos = len; + bool regex = false; + + if (TYPE(sep) == T_REGEXP) { + pos = rb_reg_search(sep, str, pos, true); + regex = true; + } + else { + StringValue(sep); + pos = str_index_for_string(RSTR(str), str_need_string(sep), + 0, -1, true, true); + } + if (pos < 0) { +failed: + return rb_ary_new3(3, rb_str_new(NULL, 0), rb_str_new(NULL,0), str); + } + if (regex) { + sep = rb_reg_nth_match(0, rb_backref_get()); + if (sep == Qnil) { + goto failed; + } + } + const long seplen = rb_str_chars_len(sep); + return rb_ary_new3(3, rstr_substr(str, 0, pos), sep, + rstr_substr(str, pos + seplen, len - pos - seplen)); +} + // NSString primitives. static void @@ -4991,6 +5080,8 @@ rb_objc_define_method(rb_cRubyString, "squeeze!", rstr_squeeze_bang, -1); rb_objc_define_method(rb_cRubyString, "sum", rstr_sum, -1); rb_objc_define_method(rb_cRubyString, "hash", rstr_hash, 0); + rb_objc_define_method(rb_cRubyString, "partition", rstr_partition, 1); + rb_objc_define_method(rb_cRubyString, "rpartition", rstr_rpartition, 1); // MacRuby extensions. rb_objc_define_method(rb_cRubyString, "transform", rstr_transform, 1);

1 0

[3708] MacRuby/branches/icu/string.c
by source_changes＠macosforge.org 05 Mar '10

05 Mar '10

Revision: 3708 http://trac.macosforge.org/projects/ruby/changeset/3708 Author: lsansonetti(a)apple.com Date: 2010-03-05 15:26:42 -0800 (Fri, 05 Mar 2010) Log Message: ----------- fixed str_compare when one of the strings cannot go uchars, added #hash Modified Paths: -------------- MacRuby/branches/icu/string.c Modified: MacRuby/branches/icu/string.c =================================================================== --- MacRuby/branches/icu/string.c 2010-03-05 23:25:44 UTC (rev 3707) +++ MacRuby/branches/icu/string.c 2010-03-05 23:26:42 UTC (rev 3708) @@ -863,15 +863,7 @@ return -1; } - if (str_is_stored_in_uchars(self) != str_is_stored_in_uchars(str)) { - // one is in uchars and the other is in binary - if (!str_try_making_data_uchars(self) - || !str_try_making_data_uchars(str)) { - // one is in uchars but the other one can't be converted in - // uchars - return -1; - } - } + str_make_same_format(self, str); const long min_len = self->length_in_bytes < str->length_in_bytes ? self->length_in_bytes : str->length_in_bytes; @@ -905,16 +897,7 @@ return -1; } - if (str_is_stored_in_uchars(self) - != str_is_stored_in_uchars(str)) { - // one is in uchars and the other is in binary - if (!str_try_making_data_uchars(self) - || !str_try_making_data_uchars(str)) { - // one is in uchars but the other one can't be converted in - // uchars - return -1; - } - } + str_make_same_format(self, str); const long min_length = self->length_in_bytes < str->length_in_bytes ? self->length_in_bytes : str->length_in_bytes; @@ -4816,6 +4799,19 @@ return rb_int2inum(sum); } +/* + * call-seq: + * str.hash => fixnum + * + * Return a hash based on the string's length and content. + */ + +static VALUE +rstr_hash(VALUE str, SEL sel) +{ + return LONG2NUM(rb_str_hash(str)); +} + // NSString primitives. static void @@ -4994,6 +4990,7 @@ rb_objc_define_method(rb_cRubyString, "squeeze", rstr_squeeze, -1); rb_objc_define_method(rb_cRubyString, "squeeze!", rstr_squeeze_bang, -1); rb_objc_define_method(rb_cRubyString, "sum", rstr_sum, -1); + rb_objc_define_method(rb_cRubyString, "hash", rstr_hash, 0); // MacRuby extensions. rb_objc_define_method(rb_cRubyString, "transform", rstr_transform, 1);

1 0

[3707] MacRuby/branches/icu/hash.c
by source_changes＠macosforge.org 05 Mar '10

05 Mar '10

Revision: 3707 http://trac.macosforge.org/projects/ruby/changeset/3707 Author: lsansonetti(a)apple.com Date: 2010-03-05 15:25:44 -0800 (Fri, 05 Mar 2010) Log Message: ----------- optimize hash cmp/hash for strings Modified Paths: -------------- MacRuby/branches/icu/hash.c Modified: MacRuby/branches/icu/hash.c =================================================================== --- MacRuby/branches/icu/hash.c 2010-03-05 23:25:24 UTC (rev 3706) +++ MacRuby/branches/icu/hash.c 2010-03-05 23:25:44 UTC (rev 3707) @@ -100,10 +100,11 @@ return a != b; } break; + + case T_STRING: + return rb_str_cmp(a, b); } - // XXX optimize for string - return !rb_eql(a, b); } @@ -120,7 +121,7 @@ return (int)a; case T_STRING: - return CFHash((CFTypeRef)a); + return (int)rb_str_hash(a); } return (int)FIX2LONG(rb_hash(a));

1 0

[3706] MacRuby/branches/icu/dispatcher.cpp
by source_changes＠macosforge.org 05 Mar '10

05 Mar '10

Revision: 3706 http://trac.macosforge.org/projects/ruby/changeset/3706 Author: lsansonetti(a)apple.com Date: 2010-03-05 15:25:24 -0800 (Fri, 05 Mar 2010) Log Message: ----------- fixed the #== optimization for strings Modified Paths: -------------- MacRuby/branches/icu/dispatcher.cpp Modified: MacRuby/branches/icu/dispatcher.cpp =================================================================== --- MacRuby/branches/icu/dispatcher.cpp 2010-03-05 12:05:52 UTC (rev 3705) +++ MacRuby/branches/icu/dispatcher.cpp 2010-03-05 23:25:24 UTC (rev 3706) @@ -1222,22 +1222,31 @@ return self == other ? Qtrue : Qfalse; case T_STRING: + if (self == other) { + return Qtrue; + } + if (TYPE(other) != self_type) { + return Qfalse; + } + return rb_str_equal(self, other); + case T_ARRAY: - case T_HASH: if (self == other) { return Qtrue; } if (TYPE(other) != self_type) { return Qfalse; } - if (self_type == T_ARRAY) { - return rb_ary_equal(self, other); + return rb_ary_equal(self, other); + + case T_HASH: + if (self == other) { + return Qtrue; } - if (self_type == T_HASH) { - return rb_hash_equal(self, other); + if (TYPE(other) != self_type) { + return Qfalse; } - return CFEqual((CFTypeRef)self, (CFTypeRef)other) - ? Qtrue : Qfalse; + return rb_hash_equal(self, other); case T_BIGNUM: return rb_big_eq(self, other);

1 0

[3705] MacRuby/branches/icu/string.c
by source_changes＠macosforge.org 05 Mar '10

05 Mar '10

Revision: 3705 http://trac.macosforge.org/projects/ruby/changeset/3705 Author: vincent.isambart(a)gmail.com Date: 2010-03-05 04:05:52 -0800 (Fri, 05 Mar 2010) Log Message: ----------- improved str_inspect Modified Paths: -------------- MacRuby/branches/icu/string.c Modified: MacRuby/branches/icu/string.c =================================================================== --- MacRuby/branches/icu/string.c 2010-03-05 05:29:12 UTC (rev 3704) +++ MacRuby/branches/icu/string.c 2010-03-05 12:05:52 UTC (rev 3705) @@ -2391,7 +2391,7 @@ static VALUE str_inspect(rb_str_t *str, bool dump) { - const bool uchars = str_is_stored_in_uchars(str); + const bool uchars = str_try_making_data_uchars(str); const long len = uchars ? str_length(str, true) : str->length_in_bytes; @@ -2409,13 +2409,20 @@ #define GET_UCHAR(pos) \ ((uchars \ - ? str->data.uchars[pos] : (UChar)str->data.bytes[pos])) + ? str->data.uchars[pos] : (unsigned char)str->data.bytes[pos])) inspect_append(result, '"', false); for (long i = 0; i < len; i++) { const UChar c = GET_UCHAR(i); - if (iswprint(c)) { + bool print; + if (uchars) { + print = iswprint(c); + } + else { // ASCII printable characters + print = ((c >= 0x20) && (c <= 0x7E)); + } + if (print) { if (c == '"' || c == '\\') { inspect_append(result, c, true); }

1 0

[3704] MacRuby/branches/icu/symbol.c
by source_changes＠macosforge.org 05 Mar '10

05 Mar '10

Revision: 3704 http://trac.macosforge.org/projects/ruby/changeset/3704 Author: lsansonetti(a)apple.com Date: 2010-03-04 21:29:12 -0800 (Thu, 04 Mar 2010) Log Message: ----------- better detect if symbols are printable Modified Paths: -------------- MacRuby/branches/icu/symbol.c Modified: MacRuby/branches/icu/symbol.c =================================================================== --- MacRuby/branches/icu/symbol.c 2010-03-05 04:12:35 UTC (rev 3703) +++ MacRuby/branches/icu/symbol.c 2010-03-05 05:29:12 UTC (rev 3704) @@ -41,6 +41,12 @@ return sym; } +static bool +is_identchar(UChar c) +{ + return isalnum(c) || c == '_' || !isascii(c); +} + ID rb_intern_str(VALUE str) { @@ -98,11 +104,9 @@ if (pos < chars_len && !isdigit(chars[pos])) { for (; pos < chars_len; pos++) { - UChar c = chars[pos]; - if (isalnum(c) || c == '_' || !isascii(c)) { - continue; + if (!is_identchar(chars[pos])) { + break; } - break; } } if (pos < chars_len) { @@ -315,6 +319,42 @@ */ static bool +is_special_global_name(UChar *ptr, long len) +{ + if (len <= 0) { + return false; + } + + long pos = 0; + switch (ptr[pos]) { + case '~': case '*': case '$': case '?': case '!': + case '@': case '/': case '\\': case ';': case ',': + case '.': case '=': case ':': case '<': case '>': + case '\"': case '&': case '`': case '\'': case '+': case '0': + pos++; + break; + + case '-': + pos++; + if (pos < len && is_identchar(ptr[pos])) { + pos++; + } + break; + + default: + if (!isdigit(ptr[pos])) { + return false; + } + do { + pos++; + } + while (pos < len && isdigit(ptr[pos])); + break; + } + return pos == len; +} + +static bool sym_should_be_escaped(VALUE sym) { UChar *chars = NULL; @@ -322,17 +362,169 @@ bool need_free = false; rb_str_get_uchars(RSYM(sym)->str, &chars, &chars_len, &need_free); - // TODO: this is really not enough, we should mimic 1.9's - // rb_enc_symname2_p() function. + if (chars_len == 0) { + return true; + } + bool escape = false; for (long i = 0; i < chars_len; i++) { - if (!iswprint(chars[i]) - || iswspace(chars[i])) { + if (!isprint(chars[i])) { escape = true; break; } } + if (escape) { + goto bail; + } + + long pos = 0; + bool localid = false; + + switch (chars[pos]) { + case '\0': + escape = true; + break; + + case '$': + pos++; + if (pos < chars_len && is_special_global_name(&chars[pos], + chars_len - pos)) { + goto bail; + } + goto id; + + case '@': + pos++; + if (pos < chars_len && chars[pos] == '@') { + pos++; + } + goto id; + + case '<': + pos++; + if (pos < chars_len) { + if (chars[pos] == '<') { + pos++; + } + else if (chars[pos] == '=') { + pos++; + if (pos < chars_len && chars[pos] == '>') { + pos++; + } + } + } + break; + + case '>': + pos++; + if (pos < chars_len) { + if (chars[pos] == '>' || chars[pos] == '=') { + pos++; + } + } + break; + + case '=': + pos++; + if (pos == chars_len) { + escape = true; + goto bail; + } + else { + if (chars[pos] == '~') { + pos++; + } + else if (chars[pos] == '=') { + pos++; + if (pos < chars_len && chars[pos] == '=') { + pos++; + } + } + else { + escape = true; + goto bail; + } + } + break; + + case '*': + pos++; + if (pos < chars_len && chars[pos] == '*') { + pos++; + } + break; + + case '+': + case '-': + pos++; + if (pos < chars_len && chars[pos] == '@') { + pos++; + } + break; + + case '|': case '^': case '&': case '/': + case '%': case '~': case '`': + pos++; + break; + + case '[': + pos++; + if (pos < chars_len && chars[pos] != ']') { + escape = true; + goto bail; + } + pos++; + if (pos < chars_len && chars[pos] == '=') { + pos++; + } + break; + + case '!': + pos++; + if (pos == chars_len) { + goto bail; + } + else { + if (chars[pos] == '=' || chars[pos] == '~') { + pos++; + } + else { + escape = true; + goto bail; + } + } + break; + + default: + localid = !isupper(chars[pos]); + // fall through + + id: + if (pos >= chars_len + || (chars[pos] != '_' && !isalpha(chars[pos]) + && isascii(chars[pos]))) { + escape = true; + goto bail; + } + while (pos < chars_len && is_identchar(chars[pos])) { + pos++; + } + if (localid) { + if (pos < chars_len + && (chars[pos] == '!' || chars[pos] == '?' + || chars[pos] == '=')) { + pos++; + } + } + break; + } + + if (pos < chars_len) { + escape = true; + } + +bail: if (need_free) { free(chars); }

1 0

[3703] MacRuby/branches/icu/symbol.c
by source_changes＠macosforge.org 05 Mar '10

05 Mar '10

Revision: 3703 http://trac.macosforge.org/projects/ruby/changeset/3703 Author: lsansonetti(a)apple.com Date: 2010-03-04 20:12:35 -0800 (Thu, 04 Mar 2010) Log Message: ----------- better symbols sanitization Modified Paths: -------------- MacRuby/branches/icu/symbol.c Modified: MacRuby/branches/icu/symbol.c =================================================================== --- MacRuby/branches/icu/symbol.c 2010-03-05 03:40:56 UTC (rev 3702) +++ MacRuby/branches/icu/symbol.c 2010-03-05 04:12:35 UTC (rev 3703) @@ -52,26 +52,32 @@ rb_sym_t *sym = NULL; - const long chars_len = rb_str_chars_len(str); + UChar *chars = NULL; + long chars_len = 0; + bool need_free = false; + rb_str_get_uchars(str, &chars, &chars_len, &need_free); + + long pos = 0; if (chars_len > 0) { - UChar c = rb_str_get_uchar(str, 0); + UChar c = chars[0]; switch (c) { case '$': id = ID_GLOBAL; - break; + goto new_id; case '@': - if (chars_len > 1 && rb_str_get_uchar(str, 1) == '@') { + if (chars_len > 1 && chars[1] == '@') { + pos++; id = ID_CLASS; } else { id = ID_INSTANCE; } + pos++; break; default: - if (chars_len > 1 - && rb_str_get_uchar(str, chars_len - 1) == '=') { + if (chars_len > 1 && chars[chars_len - 1] == '=') { // Attribute assignment. id = rb_intern_str(rb_str_substr(str, 0, chars_len - 1)); if (!is_attrset_id(id)) { @@ -90,6 +96,20 @@ } } + if (pos < chars_len && !isdigit(chars[pos])) { + for (; pos < chars_len; pos++) { + UChar c = chars[pos]; + if (isalnum(c) || c == '_' || !isascii(c)) { + continue; + } + break; + } + } + if (pos < chars_len) { + id = ID_JUNK; + } + +new_id: id |= ++last_id << ID_SCOPE_SHIFT; id_register: @@ -98,6 +118,9 @@ CFDictionarySetValue(sym_id, (const void *)name_hash, (const void *)id); CFDictionarySetValue(id_str, (const void *)id, (const void *)sym); + if (need_free) { + free(chars); + } return id; }

1 0

[3702] MacRuby/branches/icu/pack.c
by source_changes＠macosforge.org 05 Mar '10

05 Mar '10

Revision: 3702 http://trac.macosforge.org/projects/ruby/changeset/3702 Author: lsansonetti(a)apple.com Date: 2010-03-04 19:40:56 -0800 (Thu, 04 Mar 2010) Log Message: ----------- fix some problems in #pack Modified Paths: -------------- MacRuby/branches/icu/pack.c Modified: MacRuby/branches/icu/pack.c =================================================================== --- MacRuby/branches/icu/pack.c 2010-03-05 02:21:32 UTC (rev 3701) +++ MacRuby/branches/icu/pack.c 2010-03-05 03:40:56 UTC (rev 3702) @@ -63,16 +63,16 @@ unsigned char *s, *t; \ int i; \ \ - zp = xmalloc(sizeof(xtype)); \ + zp = malloc(sizeof(xtype)); \ *zp = z; \ s = (unsigned char*)zp; \ - t = xmalloc(sizeof(xtype)); \ + t = malloc(sizeof(xtype)); \ for (i=0; i<sizeof(xtype); i++) { \ t[sizeof(xtype)-i-1] = s[i]; \ } \ r = *(xtype *)t; \ - xfree(t); \ - xfree(zp); \ + free(t); \ + free(zp); \ return r; \ } @@ -366,8 +366,8 @@ #endif static const char toofew[] = "too few arguments"; -static void encodes(CFMutableDataRef,const char*,long,int,int); -static void qpencode(CFMutableDataRef,VALUE,long); +static void encodes(VALUE,const char*,long,int,int); +static void qpencode(VALUE,VALUE,long); static unsigned long utf8_to_uv(const char*,long*); @@ -454,8 +454,7 @@ p = RSTRING_PTR(fmt); pend = p + RSTRING_LEN(fmt); - CFMutableDataRef data = CFDataCreateMutable(NULL, 0); - CFMakeCollectable(data); + VALUE data = bstr_new(); items = RARRAY_LEN(ary); idx = 0; @@ -531,19 +530,19 @@ case 'A': /* arbitrary binary string (ASCII space padded) */ case 'Z': /* null terminated string */ if (plen >= len) { - CFDataAppendBytes(data, (const UInt8 *)ptr, len); + bstr_concat(data, (const UInt8 *)ptr, len); if (p[-1] == '*' && type == 'Z') { - CFDataAppendBytes(data, (const UInt8 *)nul10, 1); + bstr_concat(data, (const UInt8 *)nul10, 1); } } else { - CFDataAppendBytes(data, (const UInt8 *)ptr, plen); + bstr_concat(data, (const UInt8 *)ptr, plen); len -= plen; while (len >= 10) { - CFDataAppendBytes(data, (const UInt8 *)((type == 'A')?spc10:nul10), 10); + bstr_concat(data, (const UInt8 *)((type == 'A')?spc10:nul10), 10); len -= 10; } - CFDataAppendBytes(data, (const UInt8 *)((type == 'A')?spc10:nul10), len); + bstr_concat(data, (const UInt8 *)((type == 'A')?spc10:nul10), len); } break; @@ -563,7 +562,7 @@ byte >>= 1; else { char c = byte & 0xff; - CFDataAppendBytes(data, (const UInt8 *)&c, 1); + bstr_concat(data, (const UInt8 *)&c, 1); byte = 0; } } @@ -571,7 +570,7 @@ char c; byte >>= 7 - (len & 7); c = byte & 0xff; - CFDataAppendBytes(data, (const UInt8 *)&c, 1); + bstr_concat(data, (const UInt8 *)&c, 1); } len = j; goto grow; @@ -593,7 +592,7 @@ byte <<= 1; else { char c = byte & 0xff; - CFDataAppendBytes(data, (const UInt8 *)&c, 1); + bstr_concat(data, (const UInt8 *)&c, 1); byte = 0; } } @@ -601,7 +600,7 @@ char c; byte <<= 7 - (len & 7); c = byte & 0xff; - CFDataAppendBytes(data, (const UInt8 *)&c, 1); + bstr_concat(data, (const UInt8 *)&c, 1); } len = j; goto grow; @@ -626,13 +625,13 @@ byte >>= 4; else { char c = byte & 0xff; - CFDataAppendBytes(data, (const UInt8 *)&c, 1); + bstr_concat(data, (const UInt8 *)&c, 1); byte = 0; } } if (len & 1) { char c = byte & 0xff; - CFDataAppendBytes(data, (const UInt8 *)&c, 1); + bstr_concat(data, (const UInt8 *)&c, 1); } len = j; goto grow; @@ -657,13 +656,13 @@ byte <<= 4; else { char c = byte & 0xff; - CFDataAppendBytes(data, (const UInt8 *)&c, 1); + bstr_concat(data, (const UInt8 *)&c, 1); byte = 0; } } if (len & 1) { char c = byte & 0xff; - CFDataAppendBytes(data, (const UInt8 *)&c, 1); + bstr_concat(data, (const UInt8 *)&c, 1); } len = j; goto grow; @@ -679,7 +678,7 @@ from = NEXTFROM; c = num2i32(from); - CFDataAppendBytes(data, (const UInt8 *)&c, 1); + bstr_concat(data, (const UInt8 *)&c, 1); } break; @@ -690,7 +689,7 @@ from = NEXTFROM; s = num2i32(from); - CFDataAppendBytes(data, (const UInt8 *)OFF16(&s), NATINT_LEN(short,2)); + bstr_concat(data, (const UInt8 *)OFF16(&s), NATINT_LEN(short,2)); } break; @@ -701,7 +700,7 @@ from = NEXTFROM; i = num2i32(from); - CFDataAppendBytes(data, (const UInt8 *)OFF32(&i), NATINT_LEN(int,4)); + bstr_concat(data, (const UInt8 *)OFF32(&i), NATINT_LEN(int,4)); } break; @@ -712,7 +711,7 @@ from = NEXTFROM; l = num2i32(from); - CFDataAppendBytes(data, (const UInt8 *)OFF32(&l), NATINT_LEN(long,4)); + bstr_concat(data, (const UInt8 *)OFF32(&l), NATINT_LEN(long,4)); } break; @@ -723,7 +722,7 @@ from = NEXTFROM; rb_quad_pack(tmp, from); - CFDataAppendBytes(data, (const UInt8 *)&tmp, QUAD_SIZE); + bstr_concat(data, (const UInt8 *)&tmp, QUAD_SIZE); } break; @@ -734,7 +733,7 @@ from = NEXTFROM; s = num2i32(from); s = NATINT_HTONS(s); - CFDataAppendBytes(data, (const UInt8 *)OFF16(&s), NATINT_LEN(short,2)); + bstr_concat(data, (const UInt8 *)OFF16(&s), NATINT_LEN(short,2)); } break; @@ -745,7 +744,7 @@ from = NEXTFROM; l = num2i32(from); l = NATINT_HTONL(l); - CFDataAppendBytes(data, (const UInt8 *)OFF32(&l), NATINT_LEN(long,4)); + bstr_concat(data, (const UInt8 *)OFF32(&l), NATINT_LEN(long,4)); } break; @@ -756,7 +755,7 @@ from = NEXTFROM; s = num2i32(from); s = NATINT_HTOVS(s); - CFDataAppendBytes(data, (const UInt8 *)OFF16(&s), NATINT_LEN(short,2)); + bstr_concat(data, (const UInt8 *)OFF16(&s), NATINT_LEN(short,2)); } break; @@ -767,7 +766,7 @@ from = NEXTFROM; l = num2i32(from); l = NATINT_HTOVL(l); - CFDataAppendBytes(data, (const UInt8 *)OFF32(&l), NATINT_LEN(long,4)); + bstr_concat(data, (const UInt8 *)OFF32(&l), NATINT_LEN(long,4)); } break; @@ -778,7 +777,7 @@ from = NEXTFROM; f = RFLOAT_VALUE(rb_Float(from)); - CFDataAppendBytes(data, (const UInt8 *)&f, sizeof(float)); + bstr_concat(data, (const UInt8 *)&f, sizeof(float)); } break; @@ -790,7 +789,7 @@ from = NEXTFROM; f = RFLOAT_VALUE(rb_Float(from)); f = HTOVF(f,ftmp); - CFDataAppendBytes(data, (const UInt8 *)&f, sizeof(float)); + bstr_concat(data, (const UInt8 *)&f, sizeof(float)); } break; @@ -802,7 +801,7 @@ from = NEXTFROM; d = RFLOAT_VALUE(rb_Float(from)); d = HTOVD(d,dtmp); - CFDataAppendBytes(data, (const UInt8 *)&d, sizeof(double)); + bstr_concat(data, (const UInt8 *)&d, sizeof(double)); } break; @@ -813,7 +812,7 @@ from = NEXTFROM; d = RFLOAT_VALUE(rb_Float(from)); - CFDataAppendBytes(data, (const UInt8 *)&d, sizeof(double)); + bstr_concat(data, (const UInt8 *)&d, sizeof(double)); } break; @@ -825,7 +824,7 @@ from = NEXTFROM; f = RFLOAT_VALUE(rb_Float(from)); f = HTONF(f,ftmp); - CFDataAppendBytes(data, (const UInt8 *)&f, sizeof(float)); + bstr_concat(data, (const UInt8 *)&f, sizeof(float)); } break; @@ -837,30 +836,30 @@ from = NEXTFROM; d = RFLOAT_VALUE(rb_Float(from)); d = HTOND(d,dtmp); - CFDataAppendBytes(data, (const UInt8 *)&d, sizeof(double)); + bstr_concat(data, (const UInt8 *)&d, sizeof(double)); } break; case 'x': /* null byte */ grow: while (len >= 10) { - CFDataAppendBytes(data, (const UInt8 *)nul10, 10); + bstr_concat(data, (const UInt8 *)nul10, 10); len -= 10; } - CFDataAppendBytes(data, (const UInt8 *)nul10, len); + bstr_concat(data, (const UInt8 *)nul10, len); break; case 'X': /* back up byte */ shrink: - plen = CFDataGetLength(data); + plen = bstr_length(data); if (plen < len) { rb_raise(rb_eArgError, "X outside of string"); } - CFDataSetLength(data, plen - len); + bstr_set_length(data, plen - len); break; case '@': /* null fill to absolute position */ - len -= CFDataGetLength(data); + len -= bstr_length(data); if (len > 0) { goto grow; } @@ -887,7 +886,7 @@ rb_raise(rb_eRangeError, "pack(U): value out of range"); } le = rb_uv_to_utf8(buf, l); - CFDataAppendBytes(data, (const UInt8 *)buf, le); + bstr_concat(data, (const UInt8 *)buf, le); } break; @@ -958,14 +957,14 @@ } rb_ary_push(associates, from); rb_obj_taint(from); - CFDataAppendBytes(data, (const UInt8 *)&t, sizeof(char*)); + bstr_concat(data, (const UInt8 *)&t, sizeof(char*)); } break; case 'w': /* BER compressed integer */ while (len-- > 0) { unsigned long ul; - CFMutableDataRef bufdata = CFDataCreateMutable(NULL, 0); + VALUE bufdata = bstr_new(); char c, *bufs, *bufe; from = NEXTFROM; @@ -974,7 +973,7 @@ while (TYPE(from) == T_BIGNUM) { from = rb_big_divmod(from, big128); c = NUM2INT(RARRAY_AT(from, 1)) | 0x80; /* mod */ - CFDataAppendBytes(bufdata, (const UInt8 *)&c, sizeof(char)); + bstr_concat(bufdata, (const UInt8 *)&c, sizeof(char)); from = RARRAY_AT(from, 0); /* div */ } } @@ -987,28 +986,26 @@ while (ul) { c = ((ul & 0x7f) | 0x80); - CFDataAppendBytes(bufdata, (const UInt8 *)&c, sizeof(char)); + bstr_concat(bufdata, (const UInt8 *)&c, sizeof(char)); ul >>= 7; } - if (CFDataGetLength(bufdata) > 0) { - UInt8 *buf_beg = CFDataGetMutableBytePtr(bufdata); + if (bstr_length(bufdata) > 0) { + UInt8 *buf_beg = bstr_bytes(bufdata); bufs = (char *)buf_beg; - bufe = bufs + CFDataGetLength(bufdata) - 1; + bufe = bufs + bstr_length(bufdata) - 1; *bufs &= 0x7f; /* clear continue bit */ while (bufs < bufe) { /* reverse */ c = *bufs; *bufs++ = *bufe; *bufe-- = c; } - CFDataAppendBytes(data, buf_beg, CFDataGetLength(bufdata)); + bstr_concat(data, buf_beg, bstr_length(bufdata)); } else { c = 0; - CFDataAppendBytes(data, (const UInt8 *)&c, sizeof(char)); + bstr_concat(data, (const UInt8 *)&c, sizeof(char)); } - - CFRelease(bufdata); } break; @@ -1017,22 +1014,19 @@ } } - VALUE bres = bstr_new_with_data(CFDataGetBytePtr(data), - CFDataGetLength(data)); - // Taint the ByteString accordingly. if (OBJ_TAINTED(fmt)) { - OBJ_TAINT(bres); + OBJ_TAINT(data); } else { for (long i = 0; i < items; i++) { if (OBJ_TAINTED(RARRAY_AT(ary, i))) { - OBJ_TAINT(bres); + OBJ_TAINT(data); break; } } } - return bres; + return data; } static const char uu_table[] = @@ -1041,9 +1035,9 @@ "ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789+/"; static void -encodes(CFMutableDataRef data, const char *s, long len, int type, int tail_lf) +encodes(VALUE data, const char *s, long len, int type, int tail_lf) { - char *buff = ALLOCA_N(char, len * 4 / 3 + 6); + char *buff = (char *)malloc(len * 4 / 3 + 6); long i = 0; const char *trans = type == 'u' ? uu_table : b64_table; int padding; @@ -1065,7 +1059,7 @@ len -= 3; } if (sizeof(buff) - i < 4) { - CFDataAppendBytes(data, (const UInt8 *)buff, i); + bstr_concat(data, (const UInt8 *)buff, i); i = 0; } } @@ -1084,12 +1078,13 @@ if (tail_lf) { buff[i++] = '\n'; } - CFDataAppendBytes(data, (const UInt8 *)buff, i); + bstr_concat(data, (const UInt8 *)buff, i); + free(buff); } static const char hex_table[] = "0123456789ABCDEF"; static void -qpencode(CFMutableDataRef data, VALUE from, long len) +qpencode(VALUE data, VALUE from, long len) { char buff[1024]; long i = 0, n = 0, prev = EOF; @@ -1127,7 +1122,7 @@ prev = '\n'; } if (i > 1024 - 5) { - CFDataAppendBytes(data, (const UInt8 *)buff, i); + bstr_concat(data, (const UInt8 *)buff, i); i = 0; } s++; @@ -1137,7 +1132,7 @@ buff[i++] = '\n'; } if (i > 0) { - CFDataAppendBytes(data, (const UInt8 *)buff, i); + bstr_concat(data, (const UInt8 *)buff, i); } }

1 0

[3701] MacRuby/branches/icu/string.c
by source_changes＠macosforge.org 05 Mar '10

05 Mar '10

Revision: 3701 http://trac.macosforge.org/projects/ruby/changeset/3701 Author: lsansonetti(a)apple.com Date: 2010-03-04 18:21:32 -0800 (Thu, 04 Mar 2010) Log Message: ----------- added #bytes Modified Paths: -------------- MacRuby/branches/icu/string.c Modified: MacRuby/branches/icu/string.c =================================================================== --- MacRuby/branches/icu/string.c 2010-03-05 01:51:40 UTC (rev 3700) +++ MacRuby/branches/icu/string.c 2010-03-05 02:21:32 UTC (rev 3701) @@ -4159,7 +4159,33 @@ } /* + * Document-method: each_byte * call-seq: + * str.each_byte {|fixnum| block } => str + * + * Passes each byte in str to the given block. + * + * "hello".each_byte {|c| print c, ' ' } + * + * produces: + * + * 104 101 108 108 111 + */ + +static VALUE +rstr_each_byte(VALUE str, SEL sel) +{ + RETURN_ENUMERATOR(str, 0, 0); + + for (long i = 0; i < RSTR(str)->length_in_bytes; i++) { + rb_yield(INT2FIX(RSTR(str)->data.bytes[i])); + RETURN_IF_BROKEN(); + } + return str; +} + +/* + * call-seq: * str.succ => new_str * str.next => new_str * @@ -4946,6 +4972,8 @@ rb_objc_define_method(rb_cRubyString, "each_line", rstr_each_line, -1); rb_objc_define_method(rb_cRubyString, "chars", rstr_each_char, 0); rb_objc_define_method(rb_cRubyString, "each_char", rstr_each_char, 0); + rb_objc_define_method(rb_cRubyString, "bytes", rstr_each_byte, 0); + rb_objc_define_method(rb_cRubyString, "each_byte", rstr_each_byte, 0); rb_objc_define_method(rb_cRubyString, "succ", rstr_succ, 0); rb_objc_define_method(rb_cRubyString, "succ!", rstr_succ_bang, 0); rb_objc_define_method(rb_cRubyString, "next", rstr_succ, 0);

1 0

[3700] MacRuby/branches/icu
by source_changes＠macosforge.org 05 Mar '10

05 Mar '10

Revision: 3700 http://trac.macosforge.org/projects/ruby/changeset/3700 Author: lsansonetti(a)apple.com Date: 2010-03-04 17:51:40 -0800 (Thu, 04 Mar 2010) Log Message: ----------- fixed File.join Modified Paths: -------------- MacRuby/branches/icu/file.c MacRuby/branches/icu/include/ruby/intern.h MacRuby/branches/icu/string.c Modified: MacRuby/branches/icu/file.c =================================================================== --- MacRuby/branches/icu/file.c 2010-03-05 00:47:47 UTC (rev 3699) +++ MacRuby/branches/icu/file.c 2010-03-05 01:51:40 UTC (rev 3700) @@ -19,6 +19,7 @@ #include "dln.h" #include "objc.h" #include "vm.h" +#include "encoding.h" #ifdef HAVE_UNISTD_H #include <unistd.h> @@ -2567,7 +2568,8 @@ rb_file_s_split(VALUE klass, SEL sel, VALUE path) { FilePathStringValue(path); /* get rid of converting twice */ - return rb_assoc_new(rb_file_s_dirname(Qnil, 0, path), rb_file_s_basename(0,0,1,&path)); + return rb_assoc_new(rb_file_s_dirname(Qnil, 0, path), + rb_file_s_basename(0,0,1,&path)); } static VALUE separator; @@ -2575,46 +2577,41 @@ static VALUE rb_file_join(VALUE ary, VALUE sep) { + assert(rb_str_chars_len(sep) == 1); + UChar sep_char = rb_str_get_uchar(sep, 0); VALUE res = rb_str_new(NULL, 0); - const long count = RARRAY_LEN(ary); - if (count > 0) { - for (long i = 0; i < count; i++) { - VALUE tmp = RARRAY_AT(ary, i); - switch (TYPE(tmp)) { - case T_STRING: - break; + for (long i = 0, count = RARRAY_LEN(ary); i < count; i++) { + VALUE tmp = RARRAY_AT(ary, i); + switch (TYPE(tmp)) { + case T_STRING: + break; - case T_ARRAY: - tmp = rb_file_join(tmp, sep); - break; + case T_ARRAY: + tmp = rb_file_join(tmp, sep); + break; - default: - FilePathStringValue(tmp); - } + default: + FilePathStringValue(tmp); + } - if (i > 0 && !NIL_P(sep)) { -#if 0 -// TODO: we should probably mimic what 1.9 does here instead of this - if (CFStringHasSuffix(res, sep_cf)) { - if (CFStringHasPrefix(tmp_cf, sep_cf)) { - // Remove trailing slash from res if tmp starts with a - // slash. - CFStringDelete(res, - CFRangeMake(CFStringGetLength(res) - 1, 1)); - } + if (i > 0 && !NIL_P(sep)) { + const long res_len = rb_str_chars_len(res); + const long tmp_len = rb_str_chars_len(tmp); + + if (res_len > 0 + && rb_str_get_uchar(res, res_len - 1) == sep_char) { + if (tmp_len > 0 && rb_str_get_uchar(tmp, 0) == sep_char) { + rb_str_delete(res, res_len - 1, 1); } - else if (!CFStringHasPrefix(tmp_cf, sep_cf)) { - CFStringAppend(res, sep_cf); - } -#endif - rb_str_concat(res, sep); } - - rb_str_concat(res, tmp); + else if (tmp_len == 0 + || rb_str_get_uchar(tmp, 0) != sep_char) { + rb_str_concat(res, sep); + } } + rb_str_concat(res, tmp); } - return res; } Modified: MacRuby/branches/icu/include/ruby/intern.h =================================================================== --- MacRuby/branches/icu/include/ruby/intern.h 2010-03-05 00:47:47 UTC (rev 3699) +++ MacRuby/branches/icu/include/ruby/intern.h 2010-03-05 01:51:40 UTC (rev 3700) @@ -595,6 +595,7 @@ int rb_str_casecmp(VALUE, VALUE); VALUE rb_str_equal(VALUE str1, VALUE str2); void rb_str_update(VALUE, long, long, VALUE); +void rb_str_delete(VALUE str, long beg, long len); VALUE rb_str_split(VALUE, const char*); void rb_str_associate(VALUE, VALUE); VALUE rb_str_associated(VALUE); Modified: MacRuby/branches/icu/string.c =================================================================== --- MacRuby/branches/icu/string.c 2010-03-05 00:47:47 UTC (rev 3699) +++ MacRuby/branches/icu/string.c 2010-03-05 01:51:40 UTC (rev 3700) @@ -5614,6 +5614,17 @@ } } +void +rb_str_delete(VALUE str, long beg, long len) +{ + if (IS_RSTR(str)) { + str_delete(RSTR(str), beg, len, false); + } + else { + abort(); // TODO + } +} + int rb_str_cmp(VALUE str1, VALUE str2) {

1 0