u 0.5.0

data/ext/encoding/character/utf-8/unicode.c

@@ -0,0 +1,319 @@
+/*
+ * contents: Unicode class.
+ *
+ * Copyright © 2005 Nikolai Weibull <work@rawuncut.elitemail.org>
+ */
+
+
+#include <ruby.h>
+#include <re.h>
+#include <stdbool.h>
+#include <stddef.h>
+#include <stdint.h>
+#include <limits.h>
+#include "unicode.h"
+#include "private.h"
+#include "rb_private.h"
+#include "rb_methods.h"
+
+static VALUE mUTF8Methods;
+
+void
+need_at_least_n_arguments(int argc, int n)
+{
+        static const char *const words[] = {
+                NULL,   NULL, "two",   "three", "four",
+                "five", "six", "seven", "eight", "nine"
+        };
+
+        if (argc >= n)
+                return;
+
+        if (n == 1)
+                rb_raise(rb_eArgError, "need at least one argument");
+        else if (1 < n && n < 10)
+                rb_raise(rb_eArgError, "need at least %s arguments", words[n]);
+        else
+                rb_raise(rb_eArgError, "need at least %d arguments", n);
+}
+
+unichar
+_utf_char_validated(char const *const str, char const *const str_end)
+{
+        unichar c = utf_char_validated_n(str, str_end - str);
+        switch (c) {
+        case UTF_BAD_INPUT_UNICHAR:
+                rb_raise(rb_eArgError, "input isn’t valid UTF-8");
+        case UTF_INCOMPLETE_INPUT_UNICHAR:
+                rb_raise(rb_eArgError,
+                         "input contains an incomplete UTF-8-encoded character");
+        default:
+                return c;
+        }
+}
+
+/* TODO: instead of ‘end’, perhaps use a len/max-type parameter? */
+char *
+_utf_offset_to_pointer_validated_impl(const char *str, long offset,
+                                      const char *limit, bool noisy)
+{
+	const char *p = str;
+        long saved_offset = offset;
+
+        if (offset > 0) {
+                while (p < limit && offset-- > 0)
+                        p = utf_next(p);
+
+                if (offset > 0) {
+                        if (noisy)
+                                rb_raise(rb_eIndexError,
+                                         "index %ld lays beyond end of string",
+                                         saved_offset);
+                        else
+                                return NULL;
+                }
+        } else {
+                while (offset != 0) {
+                        const char *base = p;
+                        p += offset;
+                        while (p >= limit && (*p & 0xc0) == 0x80)
+                                p--;
+
+                        if (p < limit) {
+                                if (noisy)
+                                        rb_raise(rb_eIndexError,
+                                                "index %ld lays before beginning of string",
+                                                saved_offset);
+                                else
+                                        return NULL;
+                        }
+         
+                        offset += utf_pointer_to_offset(p, base);
+                 }
+        }
+
+	return (char *)p;
+}
+
+char *
+_utf_offset_to_pointer_validated(const char *str, long offset, const char *end)
+{
+        return _utf_offset_to_pointer_validated_impl(str, offset, end, true);
+}
+
+char *
+_utf_offset_to_pointer_failable(const char *str, long offset, const char *end)
+{
+        return _utf_offset_to_pointer_validated_impl(str, offset, end, false);
+}
+
+static char *
+rb_utf_begin_setup(VALUE str, long offset, char **base_limit, char **limit)
+{
+        char *base = RSTRING(str)->ptr;
+
+        *base_limit = RSTRING(str)->ptr + RSTRING(str)->len;
+        *limit = *base_limit;
+
+        if (offset < 0) {
+                char *tmp = base;
+                base = *base_limit;
+                *base_limit = tmp;
+        }
+
+        return base;
+}
+
+bool
+rb_utf_begin_from_offset(VALUE str, long offset, char **begin, char **limit)
+{
+        char *base_limit;
+        char *base = rb_utf_begin_setup(str, offset, &base_limit, limit);
+
+        *begin = _utf_offset_to_pointer_failable(base, offset, base_limit);
+
+        return (*begin != NULL);
+}
+
+void
+rb_utf_begin_from_offset_validated(VALUE str, long offset, char **begin,
+                                   char **limit)
+{
+        char *base_limit;
+        char *base = rb_utf_begin_setup(str, offset, &base_limit, limit);
+
+        *begin = _utf_offset_to_pointer_validated(base, offset, base_limit);
+}
+
+char *
+rb_utf_prev_validated(const char *begin, const char *p)
+{
+        char *prev = utf_find_prev(begin, p);
+        if (prev == NULL)
+                rb_raise(rb_eArgError, "input isn’t valid UTF-8");
+        return prev;
+}
+
+char *
+rb_utf_next_validated(const char *p, const char *end)
+{
+        char *next = (char *)utf_next(p);
+        if (next > end)
+                rb_raise(rb_eArgError, "input isn’t valid UTF-8");
+        return next;
+}
+
+VALUE
+rb_utf_update(VALUE str, long offset, long len, VALUE replacement)
+{
+        if (len < 0)
+                rb_raise(rb_eIndexError, "negative length %ld", len);
+
+        char *begin, *limit;
+        rb_utf_begin_from_offset_validated(str, offset, &begin, &limit);
+        char *end = _utf_offset_to_pointer_failable(begin, len, limit);
+        if (end == NULL)
+                end = limit;
+
+        rb_str_update(str, begin - RSTRING(str)->ptr, end - begin, replacement);
+
+        return replacement;
+}
+
+VALUE
+rb_utf_new(const char *str, long len)
+{
+        VALUE rbstr = rb_str_new(str, len);
+        rb_extend_object(rbstr, mUTF8Methods);
+        return rbstr;
+}
+
+VALUE
+rb_utf_new2(const char *str)
+{
+        VALUE rbstr = rb_str_new2(str);
+        rb_extend_object(rbstr, mUTF8Methods);
+        return rbstr;
+}
+
+VALUE
+rb_utf_new5(VALUE obj, const char *str, long len)
+{
+        VALUE rbstr = rb_str_new5(obj, str, len);
+        rb_extend_object(rbstr, mUTF8Methods);
+        return rbstr;
+}
+
+VALUE
+rb_utf_alloc_using(char *str)
+{
+        VALUE rbstr = rb_utf_new(NULL, 0);
+        long len = strlen(str);
+
+        RSTRING(rbstr)->ptr = str;
+        RSTRING(rbstr)->aux.capa = len;
+        RSTRING(rbstr)->len = len;
+        RSTRING(rbstr)->ptr[len] = '\0';
+
+        return rbstr;
+}
+
+VALUE
+rb_utf_dup(VALUE str)
+{
+        str = rb_str_dup(str);
+        rb_extend_object(str, mUTF8Methods);
+        return str;
+}
+
+/* TODO: rewrite this using the new offset-calculating functions. */
+long
+rb_utf_index(VALUE str, VALUE sub, long offset)
+{
+        long n_chars = utf_length_n(RSTRING(str)->ptr, RSTRING(str)->len);
+
+        if (offset < 0) {
+                offset += n_chars;
+
+                if (offset < 0)
+                        return -1;
+        }
+
+        if (n_chars - offset < utf_length(RSTRING(sub)->ptr))
+                return -1;
+
+        if (RSTRING(sub)->len == 0)
+                return offset;
+
+        char *begin = utf_offset_to_pointer(RSTRING(str)->ptr, offset);
+        long pos = rb_memsearch(RSTRING(sub)->ptr, RSTRING(sub)->len,
+                                begin, RSTRING(str)->len - (begin - RSTRING(str)->ptr));
+
+        if (pos < 0)
+                return -1;
+
+        return offset + utf_pointer_to_offset(begin, begin + pos);
+}
+
+long
+rb_utf_index_regexp(VALUE str, const char *s, const char *end, VALUE sub,
+                    long offset, bool reverse)
+{
+        long byte_offset = _utf_offset_to_pointer_validated(s, offset, end) - s;
+        long byte_startpos = rb_reg_adjust_startpos(sub, str, byte_offset, reverse);
+        long byte_index = rb_reg_search(sub, str, byte_startpos, reverse);
+        if (byte_index == -1)
+                return -1;
+        return utf_pointer_to_offset(s, s + byte_index);
+}
+
+void Init_utf8(void);
+void
+Init_utf8(void)
+{
+        VALUE mEncoding = rb_define_module("Encoding");
+        VALUE mCharacter = rb_define_module_under(mEncoding, "Character");
+        VALUE mUTF8 = rb_define_module_under(mCharacter, "UTF8");
+
+        mUTF8Methods = rb_define_module_under(mUTF8, "Methods");
+
+        rb_define_module_function(mUTF8, "collate", rb_utf_collate, 2);
+        rb_define_module_function(mUTF8, "aref", rb_utf_aref_m, -1);
+        rb_define_module_function(mUTF8, "aset", rb_utf_aset_m, -1);
+        rb_define_module_function(mUTF8, "casecmp", rb_utf_casecmp, 2);
+        rb_define_module_function(mUTF8, "center", rb_utf_center, -1);
+        rb_define_module_function(mUTF8, "chomp", rb_utf_chomp, -1);
+        rb_define_module_function(mUTF8, "chomp!", rb_utf_chomp_bang, -1);
+        rb_define_module_function(mUTF8, "chop", rb_utf_chop, 1);
+        rb_define_module_function(mUTF8, "chop!", rb_utf_chop_bang, 1);
+        rb_define_module_function(mUTF8, "count", rb_utf_count, -1);
+        rb_define_module_function(mUTF8, "delete", rb_utf_delete, -1);
+        rb_define_module_function(mUTF8, "delete!", rb_utf_delete_bang, -1);
+        rb_define_module_function(mUTF8, "each_char", rb_utf_each_char, 1);
+        rb_define_module_function(mUTF8, "index", rb_utf_index_m, -1);
+        rb_define_module_function(mUTF8, "insert", rb_utf_insert, 3);
+        rb_define_module_function(mUTF8, "lstrip", rb_utf_lstrip, 1);
+        rb_define_module_function(mUTF8, "lstrip!", rb_utf_lstrip_bang, 1);
+        rb_define_module_function(mUTF8, "rindex", rb_utf_rindex_m, -1);
+        rb_define_module_function(mUTF8, "rstrip", rb_utf_rstrip, 1);
+        rb_define_module_function(mUTF8, "rstrip!", rb_utf_rstrip_bang, 1);
+        rb_define_module_function(mUTF8, "squeeze", rb_utf_squeeze, -1);
+        rb_define_module_function(mUTF8, "squeeze!", rb_utf_squeeze_bang, -1);
+        rb_define_module_function(mUTF8, "strip", rb_utf_strip, 1);
+        rb_define_module_function(mUTF8, "strip!", rb_utf_strip_bang, 1);
+        rb_define_module_function(mUTF8, "to_i", rb_utf_to_i, -1);
+        rb_define_module_function(mUTF8, "hex", rb_utf_hex, 1);
+        rb_define_module_function(mUTF8, "oct", rb_utf_oct, 1);
+        rb_define_module_function(mUTF8, "tr", rb_utf_tr, 3);
+        rb_define_module_function(mUTF8, "tr_s", rb_utf_tr_s, 3);
+
+        rb_define_module_function(mUTF8, "downcase", rb_utf_downcase, 1);
+        rb_define_module_function(mUTF8, "ljust", rb_utf_ljust, -1);
+        rb_define_module_function(mUTF8, "length", rb_utf_length, 1);
+        rb_define_module_function(mUTF8, "reverse", rb_utf_reverse, 1);
+        rb_define_module_function(mUTF8, "rjust", rb_utf_rjust, -1);
+        rb_define_module_function(mUTF8, "upcase", rb_utf_upcase, 1);
+
+        rb_define_module_function(mUTF8, "foldcase", rb_utf_foldcase, 1);
+        rb_define_module_function(mUTF8, "normalize", rb_utf_normalize, -1);
+}

data/ext/encoding/character/utf-8/unicode.h

@@ -0,0 +1,216 @@
+/*
+ * contents: Unicode handling.
+ *
+ * Copyright (C) 2004 Nikolai Weibull <source@pcppopper.org>
+ */
+
+
+#ifndef UNICODE_H
+#define UNICODE_H
+
+#if __GNUC__ > 2 || (__GNUC__ == 2 && __GNUC_MINOR__ > 4)
+#define CONST_FUNC                              \
+          __attribute__((__const__))
+#else
+#define CONST_FUNC
+#endif
+
+typedef uint32_t unichar;
+
+#define MAXUNICHAR UINT32_MAX
+
+#define MAX_UNICHAR_BYTE_LENGTH 6
+
+#define UNICODE_N_CODEPOINTS            (0x10ffff + 1)
+
+/* unichar return used for representing bad input to a function. */
+#define UTF_BAD_INPUT_UNICHAR		((unichar)-1)
+
+
+/* unichar return used for representing an incomplete input to a function. */
+#define UTF_INCOMPLETE_INPUT_UNICHAR	((unichar)-2)
+
+
+typedef enum {
+	UNICODE_CONTROL,
+	UNICODE_FORMAT,
+	UNICODE_UNASSIGNED,
+	UNICODE_PRIVATE_USE,
+	UNICODE_SURROGATE,
+	UNICODE_LOWERCASE_LETTER,
+	UNICODE_MODIFIER_LETTER,
+	UNICODE_OTHER_LETTER,
+	UNICODE_TITLECASE_LETTER,
+	UNICODE_UPPERCASE_LETTER,
+	UNICODE_COMBINING_MARK,
+	UNICODE_ENCLOSING_MARK,
+	UNICODE_NON_SPACING_MARK,
+	UNICODE_DECIMAL_NUMBER,
+	UNICODE_LETTER_NUMBER,
+	UNICODE_OTHER_NUMBER,
+	UNICODE_CONNECT_PUNCTUATION,
+	UNICODE_DASH_PUNCTUATION,
+	UNICODE_CLOSE_PUNCTUATION,
+	UNICODE_FINAL_PUNCTUATION,
+	UNICODE_INITIAL_PUNCTUATION,
+	UNICODE_OTHER_PUNCTUATION,
+	UNICODE_OPEN_PUNCTUATION,
+	UNICODE_CURRENCY_SYMBOL,
+	UNICODE_MODIFIER_SYMBOL,
+	UNICODE_MATH_SYMBOL,
+	UNICODE_OTHER_SYMBOL,
+	UNICODE_LINE_SEPARATOR,
+	UNICODE_PARAGRAPH_SEPARATOR,
+	UNICODE_SPACE_SEPARATOR
+} UnicodeType;
+
+bool unichar_isalnum(unichar c);
+bool unichar_isalpha(unichar c);
+bool unichar_iscntrl(unichar c);
+bool unichar_isdigit(unichar c);
+bool unichar_isgraph(unichar c);
+bool unichar_islower(unichar c);
+bool unichar_isprint(unichar c);
+bool unichar_ispunct(unichar c);
+bool unichar_isspace(unichar c);
+bool unichar_isupper(unichar c);
+bool unichar_istitle(unichar c);
+bool unichar_isnewline(unichar c);
+bool unichar_isxdigit(unichar c);
+bool unichar_isassigned(unichar c);
+bool unichar_iswide(unichar c);
+bool unichar_isvalid(unichar c);
+
+unichar unichar_toupper(unichar c);
+unichar unichar_tolower(unichar c);
+unichar unichar_totitle(unichar c);
+
+int unichar_digit_value(unichar c);
+int unichar_xdigit_value(unichar c);
+
+UnicodeType unichar_type(unichar c);
+
+int unichar_combining_class(unichar c) CONST_FUNC;
+
+bool unichar_mirror(unichar c, unichar *mirrored);
+
+
+typedef enum {
+	UNICODE_BREAK_MANDATORY,
+	UNICODE_BREAK_CARRIAGE_RETURN,
+	UNICODE_BREAK_LINE_FEED,
+	UNICODE_BREAK_COMBINING_MARK,
+	UNICODE_BREAK_SURROGATE,
+	UNICODE_BREAK_ZERO_WIDTH_SPACE,
+	UNICODE_BREAK_INSEPARABLE,
+	UNICODE_BREAK_NON_BREAKING_GLUE,
+	UNICODE_BREAK_CONTINGENT,
+	UNICODE_BREAK_SPACE,
+	UNICODE_BREAK_AFTER,
+	UNICODE_BREAK_BEFORE,
+	UNICODE_BREAK_BEFORE_AND_AFTER,
+	UNICODE_BREAK_HYPHEN,
+	UNICODE_BREAK_NON_STARTER,
+	UNICODE_BREAK_OPEN_PUNCTUATION,
+	UNICODE_BREAK_CLOSE_PUNCTUATION,
+	UNICODE_BREAK_QUOTATION,
+	UNICODE_BREAK_EXCLAMATION,
+	UNICODE_BREAK_IDEOGRAPHIC,
+	UNICODE_BREAK_NUMERIC,
+	UNICODE_BREAK_INFIX_SEPARATOR,
+	UNICODE_BREAK_SYMBOL,
+	UNICODE_BREAK_ALPHABETIC,
+	UNICODE_BREAK_PREFIX,
+	UNICODE_BREAK_POSTFIX,
+	UNICODE_BREAK_COMPLEX_CONTEXT,
+	UNICODE_BREAK_AMBIGUOUS,
+	UNICODE_BREAK_UNKNOWN,
+	UNICODE_BREAK_NEXT_LINE,
+	UNICODE_BREAK_WORD_JOINER,
+        UNICODE_BREAK_HANGUL_L_JAMO,
+        UNICODE_BREAK_HANGUL_V_JAMO,
+        UNICODE_BREAK_HANGUL_T_JAMO,
+        UNICODE_BREAK_HANGUL_LV_SYLLABLE,
+        UNICODE_BREAK_HANGUL_LVT_SYLLABLE
+} UnicodeBreakType;
+
+UnicodeBreakType unichar_break_type(unichar c);
+
+
+typedef enum {
+	NORMALIZE_DEFAULT,
+	NORMALIZE_NFD = NORMALIZE_DEFAULT,
+	NORMALIZE_DEFAULT_COMPOSE,
+	NORMALIZE_NFC = NORMALIZE_DEFAULT_COMPOSE,
+	NORMALIZE_ALL,
+	NORMALIZE_NFKD = NORMALIZE_ALL,
+	NORMALIZE_ALL_COMPOSE,
+	NORMALIZE_NFKC = NORMALIZE_ALL_COMPOSE
+} NormalizeMode;
+
+void unicode_canonical_ordering(unichar *str, size_t len);
+unichar *unicode_canonical_decomposition(unichar c, size_t *result_len);
+
+char *utf_normalize(const char *str, NormalizeMode mode);
+char *utf_normalize_n(const char *str, NormalizeMode mode, size_t len);
+
+
+
+
+char *utf_upcase(const char *str);
+char *utf_upcase_n(const char *str, size_t len);
+char *utf_downcase(const char *str);
+char *utf_downcase_n(const char *str, size_t len);
+char *utf_foldcase(const char *str);
+char *utf_foldcase_n(const char *str, size_t len);
+
+unichar utf_char(const char *str);
+unichar utf_char_n(const char *str, size_t max);
+unichar utf_char_validated(const char *str);
+unichar utf_char_validated_n(const char *str, size_t max);
+
+extern const char * const s_utf_skip_lengths;
+#define utf_next(str)	((str) + s_utf_skip_lengths[*(const unsigned char *)(str)])
+char *utf_find_next(const char *p, const char *end);
+char *utf_prev(const char *p);
+char *utf_find_prev(const char *begin, const char *p);
+char *utf_offset_to_pointer(const char *str, long offset);
+long utf_pointer_to_offset(const char *str, const char *pos);
+
+void utf_copy(char *dest, const char *src);
+void utf_copy_n(char *dest, const char *src, size_t n);
+void utf_append(char *dest, const char *src);
+void utf_append_n(char *dest, const char *src, size_t n);
+int utf_collate(const char *a, const char *b);
+char *utf_collate_key(const char *str);
+char *utf_collate_key_n(const char *str, size_t len);
+int utf_char_index(const char *str, unichar c);
+int utf_char_index_n(const char *str, unichar c, size_t len);
+int utf_char_rindex(const char *str, unichar c);
+int utf_char_rindex_n(const char *str, unichar c, size_t len);
+int utf_index(const char *haystack, const char *needle);
+int utf_index_n(const char *haystack, const char *needle, size_t len);
+int utf_rindex(const char *haystack, const char *needle);
+int utf_rindex_n(const char *haystack, const char *needle, size_t len);
+bool utf_has_prefix(const char *str, const char *prefix);
+long utf_length(const char *str);
+long utf_length_n(const char *str, long len);
+size_t utf_width(const char *str);
+size_t utf_width_n(const char *str, size_t len);
+size_t utf_byte_length(const char *str);
+char *utf_reverse(const char *str);
+char *utf_reverse_n(const char *str, size_t len);
+
+bool utf_isvalid(const char *str);
+bool utf_isvalid_n(const char *str, size_t max, const char **end);
+
+/* XXX: should probably name stuff utf32 instead of ucs4 */
+int unichar_to_utf(unichar c, char *result);
+char *ucs4_to_utf8(unichar *str, size_t *items_read, size_t *items_written);
+char *ucs4_to_utf8_n(unichar *str, size_t len, size_t *items_read, size_t *items_written);
+unichar *utf8_to_ucs4_fast(const char *str, size_t *items_written);
+unichar *utf8_to_ucs4_fast_n(const char *str, size_t len, size_t *items_written);
+unichar *utf8_to_ucs4(const char *str, size_t *items_read, size_t *items_written);
+unichar *utf8_to_ucs4_n(const char *str, int len, size_t *items_read, size_t *items_written);
+
+#endif /* UNICODE_H */