byk 0.4.0 → 1.1.0

data/ext/byk/byk.c

@@ -1,314 +1,380 @@
-#include <stdio.h>
 #include <ruby.h>
 #include <ruby/encoding.h>
 
-#ifndef rb_check_arity
-#define rb_check_arity rb_check_arity
-
-NORETURN(void rb_error_arity(int, int, int));
+#define STR_ENC_GET(str) rb_enc_from_index(ENCODING_GET(str))
 
 static inline void
-rb_check_arity(int argc, int min, int max)
+_str_cat_char(VALUE str, unsigned c, rb_encoding *enc)
 {
-    if ((argc < min) || (max != -1 && argc > max))
-	rb_error_arity(argc, min, max);
+    char s[16];
+    int n = rb_enc_codelen(c, enc);
+    rb_enc_mbcput(c, s, enc);
+    rb_str_buf_cat(str, s, n);
 }
-#endif
-
-#define STR_ENC_GET(str) rb_enc_from_index(ENCODING_GET(str))
-
-#define STR_CAT_COND_ASCII(ascii, dest, chr, ascii_chr, len, enc) \
-    ascii ? rb_enc_str_buf_cat(dest, chr, len, enc)               \
-          : str_cat_char(dest, ascii_chr, enc)
 
 enum {
-    LAT_CAP_TJ=262,
-    LAT_TJ,
-    LAT_CAP_CH=268,
-    LAT_CH,
-    LAT_CAP_DJ=272,
-    LAT_DJ,
-    LAT_CAP_SH=352,
-    LAT_SH,
-    LAT_CAP_ZH=381,
-    LAT_ZH,
-    CYR_CAP_DJ=1026,
-    CYR_CAP_J=1032,
-    CYR_CAP_LJ,
-    CYR_CAP_NJ,
-    CYR_CAP_TJ,
-    CYR_CAP_DZ=1039,
-    CYR_CAP_A,
-    CYR_CAP_B,
-    CYR_CAP_V,
-    CYR_CAP_G,
-    CYR_CAP_D,
-    CYR_CAP_E,
-    CYR_CAP_ZH,
-    CYR_CAP_Z,
-    CYR_CAP_I,
-    CYR_CAP_K=1050,
-    CYR_CAP_L,
-    CYR_CAP_M,
-    CYR_CAP_N,
-    CYR_CAP_O,
-    CYR_CAP_P,
-    CYR_CAP_R,
-    CYR_CAP_S,
-    CYR_CAP_T,
-    CYR_CAP_U,
-    CYR_CAP_F,
-    CYR_CAP_H,
-    CYR_CAP_C,
-    CYR_CAP_CH,
-    CYR_CAP_SH,
-    CYR_A=1072,
-    CYR_B,
-    CYR_V,
-    CYR_G,
-    CYR_D,
-    CYR_E,
-    CYR_ZH,
-    CYR_Z,
-    CYR_I,
-    CYR_K=1082,
-    CYR_L,
-    CYR_M,
-    CYR_N,
-    CYR_O,
-    CYR_P,
-    CYR_R,
-    CYR_S,
-    CYR_T,
-    CYR_U,
-    CYR_F,
-    CYR_H,
-    CYR_C,
-    CYR_CH,
-    CYR_SH,
-    CYR_DJ=1106,
-    CYR_J=1112,
-    CYR_LJ,
-    CYR_NJ,
-    CYR_TJ,
-    CYR_DZ=1119
+    LAT_CAP_TJ=262, LAT_TJ, LAT_CAP_CH=268, LAT_CH,
+    LAT_CAP_DJ=272, LAT_DJ, LAT_CAP_SH=352, LAT_SH,
+    LAT_CAP_ZH=381, LAT_ZH, CYR_CAP_DJ=1026, CYR_CAP_J=1032,
+    CYR_CAP_LJ, CYR_CAP_NJ, CYR_CAP_TJ, CYR_CAP_DZ=1039,
+    CYR_CAP_A, CYR_CAP_B, CYR_CAP_V, CYR_CAP_G,
+    CYR_CAP_D, CYR_CAP_E, CYR_CAP_ZH, CYR_CAP_Z,
+    CYR_CAP_I, CYR_CAP_K=1050, CYR_CAP_L, CYR_CAP_M,
+    CYR_CAP_N, CYR_CAP_O, CYR_CAP_P, CYR_CAP_R,
+    CYR_CAP_S, CYR_CAP_T, CYR_CAP_U, CYR_CAP_F,
+    CYR_CAP_H, CYR_CAP_C, CYR_CAP_CH, CYR_CAP_SH,
+    CYR_A=1072, CYR_B, CYR_V, CYR_G, CYR_D,
+    CYR_E, CYR_ZH, CYR_Z, CYR_I, CYR_K=1082,
+    CYR_L, CYR_M, CYR_N, CYR_O, CYR_P,
+    CYR_R, CYR_S, CYR_T, CYR_U, CYR_F,
+    CYR_H, CYR_C, CYR_CH, CYR_SH, CYR_DJ=1106,
+    CYR_J=1112, CYR_LJ, CYR_NJ, CYR_TJ, CYR_DZ=1119
 };
 
-static inline unsigned int
-is_upper_case(unsigned int c)
+static inline unsigned
+is_cap(unsigned codepoint)
 {
-    return ((c >= 65 && c <= 90)
-            || (c >= CYR_CAP_DJ && c <= CYR_CAP_SH)
-            || c == LAT_CAP_TJ
-            || c == LAT_CAP_CH
-            || c == LAT_CAP_DJ
-            || c == LAT_CAP_SH
-            || c == LAT_CAP_ZH);
+    if (codepoint >= 65 && codepoint <= 90) return 1;
+    if (codepoint >= CYR_CAP_DJ && codepoint <= CYR_CAP_SH) return 1;
+
+    switch(codepoint) {
+    case LAT_CAP_TJ:
+    case LAT_CAP_CH:
+    case LAT_CAP_DJ:
+    case LAT_CAP_SH:
+    case LAT_CAP_ZH:
+        return 1;
+    default:
+        return 0;
+    }
 }
 
-static void
-str_cat_char(VALUE str, unsigned int c, rb_encoding *enc)
+static inline unsigned
+is_digraph(unsigned codepoint)
 {
-    char s[16];
-    int n = rb_enc_codelen(c, enc);
-    rb_enc_mbcput(c, s, enc);
-    rb_enc_str_buf_cat(str, s, n, enc);
+    switch(codepoint) {
+    case CYR_LJ:
+    case CYR_NJ:
+    case CYR_DZ:
+    case CYR_CAP_LJ:
+    case CYR_CAP_NJ:
+    case CYR_CAP_DZ:
+        return 1;
+    default:
+        return 0;
+    }
+}
+
+static unsigned
+digraph_to_cyr(unsigned codepoint, unsigned codepoint2, unsigned capitalize, unsigned *next_out)
+{
+    static unsigned CYR_MAP[] = {
+        CYR_A, CYR_B, CYR_C, CYR_D, CYR_E, CYR_F,
+        CYR_G, CYR_H, CYR_I, CYR_J, CYR_K, CYR_L,
+        CYR_M, CYR_N, CYR_O, CYR_P, 0, CYR_R,
+        CYR_S, CYR_T, CYR_U, CYR_V, 0, 0, 0, CYR_Z
+    };
+
+    static unsigned CYR_CAPS_MAP[] = {
+        CYR_CAP_A, CYR_CAP_B, CYR_CAP_C, CYR_CAP_D, CYR_CAP_E, CYR_CAP_F,
+        CYR_CAP_G, CYR_CAP_H, CYR_CAP_I, CYR_CAP_J, CYR_CAP_K, CYR_CAP_L,
+        CYR_CAP_M, CYR_CAP_N, CYR_CAP_O, CYR_CAP_P, 0, CYR_CAP_R,
+        CYR_CAP_S, CYR_CAP_T, CYR_CAP_U, CYR_CAP_V, 0, 0, 0, CYR_CAP_Z
+    };
+
+    if (codepoint2 == LAT_CAP_ZH || codepoint2 == LAT_ZH) {
+        switch (codepoint) {
+        case 'd': return CYR_DZ;
+        case 'D': return CYR_CAP_DZ;
+        }
+    }
+
+    if (codepoint2 == 'j' || codepoint2 == 'J') {
+        switch (codepoint) {
+        case 'l': return CYR_LJ;
+        case 'n': return CYR_NJ;
+        case 'L': return CYR_CAP_LJ;
+        case 'N': return CYR_CAP_NJ;
+        }
+    }
+
+    if (codepoint >= 'a' && codepoint <= 'z') return CYR_MAP[codepoint - 'a'];
+    if (codepoint >= 'A' && codepoint <= 'Z') return CYR_CAPS_MAP[codepoint - 'A'];
+
+    switch (codepoint) {
+    case LAT_CH: return CYR_CH;
+    case LAT_DJ: return CYR_DJ;
+    case LAT_SH: return CYR_SH;
+    case LAT_TJ: return CYR_TJ;
+    case LAT_ZH: return CYR_ZH;
+    case LAT_CAP_CH: return CYR_CAP_CH;
+    case LAT_CAP_DJ: return CYR_CAP_DJ;
+    case LAT_CAP_SH: return CYR_CAP_SH;
+    case LAT_CAP_TJ: return CYR_CAP_TJ;
+    case LAT_CAP_ZH: return CYR_CAP_ZH;
+    }
+
+    return 0;
+}
+
+static unsigned
+digraph_to_latin(unsigned codepoint, unsigned codepoint2, unsigned capitalize, unsigned *next_out)
+{
+    static char LAT_MAP[] = {
+        'a', 'b', 'v', 'g', 'd', 'e', 0, 'z', 'i', 0, 'k', 'l',
+        'm', 'n', 'o', 'p', 'r', 's', 't', 'u', 'f', 'h', 'c'
+    };
+
+    static char LAT_CAPS_MAP[] = {
+        'A', 'B', 'V', 'G', 'D', 'E', 0, 'Z', 'I', 0, 'K', 'L',
+        'M', 'N', 'O', 'P', 'R', 'S', 'T', 'U', 'F', 'H', 'C'
+    };
+
+    if (codepoint < CYR_CAP_DJ || codepoint > CYR_DZ) return 0;
+
+    switch (codepoint) {
+    case CYR_ZH: return LAT_ZH;
+    case CYR_CAP_ZH: return LAT_CAP_ZH;
+    }
+
+    if (codepoint >= CYR_A && codepoint <= CYR_C)
+        return LAT_MAP[codepoint - CYR_A];
+
+    if (codepoint >= CYR_CAP_A && codepoint <= CYR_CAP_C)
+        return LAT_CAPS_MAP[codepoint - CYR_CAP_A];
+
+    if (codepoint >= CYR_A) {
+        switch (codepoint) {
+        case CYR_J:  return 'j';
+        case CYR_TJ: return LAT_TJ;
+        case CYR_CH: return LAT_CH;
+        case CYR_SH: return LAT_SH;
+        case CYR_DJ: return LAT_DJ;
+        case CYR_LJ: *next_out = 'j'; return 'l';
+        case CYR_NJ: *next_out = 'j'; return 'n';
+        case CYR_DZ: *next_out = LAT_ZH; return 'd';
+        }
+    }
+    else {
+        switch (codepoint) {
+        case CYR_CAP_J:  return 'J';
+        case CYR_CAP_TJ: return LAT_CAP_TJ;
+        case CYR_CAP_CH: return LAT_CAP_CH;
+        case CYR_CAP_SH: return LAT_CAP_SH;
+        case CYR_CAP_DJ: return LAT_CAP_DJ;
+        case CYR_CAP_LJ: *next_out = (capitalize || is_cap(codepoint2)) ? 'J' : 'j'; return 'L';
+        case CYR_CAP_NJ: *next_out = (capitalize || is_cap(codepoint2)) ? 'J' : 'j'; return 'N';
+        case CYR_CAP_DZ: *next_out = (capitalize || is_cap(codepoint2)) ? LAT_CAP_ZH : LAT_ZH; return 'D';
+        }
+    }
+
+    return 0;
+}
+
+static unsigned
+digraph_to_ascii(unsigned codepoint, unsigned codepoint2, unsigned capitalize, unsigned *next_out)
+{
+    switch (codepoint) {
+    case LAT_TJ:
+    case LAT_CH:
+    case CYR_TJ:
+    case CYR_CH: return 'c';
+    case LAT_SH:
+    case CYR_SH: return 's';
+    case LAT_ZH:
+    case CYR_ZH: return 'z';
+    case LAT_DJ:
+    case CYR_DJ: *next_out = 'j'; return 'd';
+    case LAT_CAP_TJ:
+    case LAT_CAP_CH:
+    case CYR_CAP_TJ:
+    case CYR_CAP_CH: return 'C';
+    case LAT_CAP_SH:
+    case CYR_CAP_SH: return 'S';
+    case LAT_CAP_ZH:
+    case CYR_CAP_ZH: return 'Z';
+    case LAT_CAP_DJ:
+    case CYR_CAP_DJ:
+        *next_out = (capitalize || is_cap(codepoint2)) ? 'J' : 'j'; return 'D';
+    case CYR_DZ:
+        *next_out = (capitalize || is_cap(codepoint2)) ? 'Z' : 'z'; return 'd';
+    case CYR_CAP_DZ:
+        *next_out = (capitalize || is_cap(codepoint2)) ? 'Z' : 'z'; return 'D';
+    default:
+        return digraph_to_latin(codepoint, codepoint2, capitalize, next_out);
+    }
 }
 
 static VALUE
-str_to_latin(int argc, VALUE *argv, VALUE str, int ascii, int bang)
+str_to_srb(VALUE str, int strategy, int bang)
 {
     VALUE dest;
-    long dest_len;
-    char *pos, *end;
     rb_encoding *enc;
+
     int len, next_len;
-    int seen_upper = 0;
-    int force_upper = 0;
-    unsigned int codepoint = 0;
-    unsigned int next_codepoint = 0;
+    unsigned in, in2, out, out2, seen_cap = 0;
+    char *pos, *end, *seq_start = 0;
 
-    rb_check_arity(argc, 0, 1);
+    unsigned (*method)(unsigned, unsigned, unsigned, unsigned*);
+
+    switch(strategy) {
+    case 0:  method = &digraph_to_cyr;   break;
+    case 1:  method = &digraph_to_latin; break;
+    default: method = &digraph_to_ascii;
+    }
 
+    StringValue(str);
     pos = RSTRING_PTR(str);
     if (!pos || RSTRING_LEN(str) == 0) return str;
 
     end = RSTRING_END(str);
     enc = STR_ENC_GET(str);
-    dest_len = RSTRING_LEN(str) + 30;
-    dest = rb_str_buf_new(dest_len);
+    dest = rb_str_buf_new(RSTRING_LEN(str) + 30);
     rb_enc_associate(dest, enc);
 
-    codepoint = rb_enc_codepoint_len(pos, end, &len, enc);
+    in = rb_enc_codepoint_len(pos, end, &len, enc);
 
     while (pos < end) {
-        if (pos + len < end) {
-            next_codepoint = rb_enc_codepoint_len(pos + len, end, &next_len, enc);
-        }
+        in2 = out2 = 0;
 
-        force_upper = seen_upper || is_upper_case(next_codepoint);
-        seen_upper = is_upper_case(codepoint);
-
-        /* Latin -> "ASCII Latin" conversion */
-        if (ascii && codepoint >= LAT_CAP_TJ && codepoint <= LAT_ZH) {
-            switch (codepoint) {
-            case LAT_TJ:
-            case LAT_CH:     rb_enc_str_buf_cat(dest, "c", 1, enc); break;
-            case LAT_DJ:     rb_enc_str_buf_cat(dest, "dj", 2, enc); break;
-            case LAT_SH:     rb_enc_str_buf_cat(dest, "s", 1, enc); break;
-            case LAT_ZH:     rb_enc_str_buf_cat(dest, "z", 1, enc); break;
-            case LAT_CAP_TJ:
-            case LAT_CAP_CH: rb_enc_str_buf_cat(dest, "C", 1, enc); break;
-            case LAT_CAP_SH: rb_enc_str_buf_cat(dest, "S", 1, enc); break;
-            case LAT_CAP_ZH: rb_enc_str_buf_cat(dest, "Z", 1, enc); break;
-
-            case LAT_CAP_DJ:
-                force_upper ? rb_enc_str_buf_cat(dest, "DJ", 2, enc)
-                            : rb_enc_str_buf_cat(dest, "Dj", 2, enc);
-                break;
-            default:
-                rb_enc_str_buf_cat(dest, pos, len, enc);
-            }
-        }
+        if (pos + len < end)
+            in2 = rb_enc_codepoint_len(pos + len, end, &next_len, enc);
 
-        /* Non-Cyrillic codepoints */
-        else if (codepoint < CYR_CAP_DJ || codepoint > CYR_DZ) {
-            rb_enc_str_buf_cat(dest, pos, len, enc);
-        }
+        out = (*method)(in, in2, seen_cap, &out2);
 
-        /* Cyrillic -> Latin conversion */
-        else if (codepoint >= CYR_A) {
-            switch (codepoint) {
-            case CYR_A:      rb_enc_str_buf_cat(dest, "a",  1, enc); break;
-            case CYR_B:      rb_enc_str_buf_cat(dest, "b",  1, enc); break;
-            case CYR_V:      rb_enc_str_buf_cat(dest, "v",  1, enc); break;
-            case CYR_G:      rb_enc_str_buf_cat(dest, "g",  1, enc); break;
-            case CYR_D:      rb_enc_str_buf_cat(dest, "d",  1, enc); break;
-            case CYR_E:      rb_enc_str_buf_cat(dest, "e",  1, enc); break;
-            case CYR_Z:      rb_enc_str_buf_cat(dest, "z",  1, enc); break;
-            case CYR_I:      rb_enc_str_buf_cat(dest, "i",  1, enc); break;
-            case CYR_K:      rb_enc_str_buf_cat(dest, "k",  1, enc); break;
-            case CYR_L:      rb_enc_str_buf_cat(dest, "l",  1, enc); break;
-            case CYR_M:      rb_enc_str_buf_cat(dest, "m",  1, enc); break;
-            case CYR_N:      rb_enc_str_buf_cat(dest, "n",  1, enc); break;
-            case CYR_O:      rb_enc_str_buf_cat(dest, "o",  1, enc); break;
-            case CYR_P:      rb_enc_str_buf_cat(dest, "p",  1, enc); break;
-            case CYR_R:      rb_enc_str_buf_cat(dest, "r",  1, enc); break;
-            case CYR_S:      rb_enc_str_buf_cat(dest, "s",  1, enc); break;
-            case CYR_T:      rb_enc_str_buf_cat(dest, "t",  1, enc); break;
-            case CYR_U:      rb_enc_str_buf_cat(dest, "u",  1, enc); break;
-            case CYR_F:      rb_enc_str_buf_cat(dest, "f",  1, enc); break;
-            case CYR_H:      rb_enc_str_buf_cat(dest, "h",  1, enc); break;
-            case CYR_C:      rb_enc_str_buf_cat(dest, "c",  1, enc); break;
-            case CYR_J:      rb_enc_str_buf_cat(dest, "j",  1, enc); break;
-            case CYR_LJ:     rb_enc_str_buf_cat(dest, "lj", 2, enc); break;
-            case CYR_NJ:     rb_enc_str_buf_cat(dest, "nj", 2, enc); break;
-            case CYR_DJ:     STR_CAT_COND_ASCII(ascii, dest, "dj", LAT_DJ, 2, enc); break;
-            case CYR_TJ:     STR_CAT_COND_ASCII(ascii, dest, "c", LAT_TJ, 1, enc); break;
-            case CYR_CH:     STR_CAT_COND_ASCII(ascii, dest, "c", LAT_CH, 1, enc); break;
-            case CYR_ZH:     STR_CAT_COND_ASCII(ascii, dest, "z", LAT_ZH, 1, enc); break;
-            case CYR_SH:     STR_CAT_COND_ASCII(ascii, dest, "s", LAT_SH, 1, enc); break;
-            case CYR_DZ:
-                rb_enc_str_buf_cat(dest, "d", 1, enc);
-                STR_CAT_COND_ASCII(ascii, dest, "z", LAT_ZH, 1, enc);
-                break;
-            default:
-                rb_enc_str_buf_cat(dest, pos, len, enc);
+        if (out) {
+            /* flush previous untranslatable sequence */
+            if (seq_start) {
+                rb_str_buf_cat(dest, seq_start, pos - seq_start);
+                seq_start = 0;
             }
+
+            _str_cat_char(dest, out, enc);
+            if (out2) _str_cat_char(dest, out2, enc);
+        }
+        else if (!seq_start) {
+            /* mark the beginning of an untranslatable sequence */
+            seq_start = pos;
         }
 
-        /* Cyrillic -> Latin conversion, caps */
-        else {
-            switch (codepoint) {
-            case CYR_CAP_J:  rb_enc_str_buf_cat(dest, "J",  1, enc); break;
-            case CYR_CAP_A:  rb_enc_str_buf_cat(dest, "A",  1, enc); break;
-            case CYR_CAP_B:  rb_enc_str_buf_cat(dest, "B",  1, enc); break;
-            case CYR_CAP_V:  rb_enc_str_buf_cat(dest, "V",  1, enc); break;
-            case CYR_CAP_G:  rb_enc_str_buf_cat(dest, "G",  1, enc); break;
-            case CYR_CAP_D:  rb_enc_str_buf_cat(dest, "D",  1, enc); break;
-            case CYR_CAP_E:  rb_enc_str_buf_cat(dest, "E",  1, enc); break;
-            case CYR_CAP_Z:  rb_enc_str_buf_cat(dest, "Z",  1, enc); break;
-            case CYR_CAP_I:  rb_enc_str_buf_cat(dest, "I",  1, enc); break;
-            case CYR_CAP_K:  rb_enc_str_buf_cat(dest, "K",  1, enc); break;
-            case CYR_CAP_L:  rb_enc_str_buf_cat(dest, "L",  1, enc); break;
-            case CYR_CAP_M:  rb_enc_str_buf_cat(dest, "M",  1, enc); break;
-            case CYR_CAP_N:  rb_enc_str_buf_cat(dest, "N",  1, enc); break;
-            case CYR_CAP_O:  rb_enc_str_buf_cat(dest, "O",  1, enc); break;
-            case CYR_CAP_P:  rb_enc_str_buf_cat(dest, "P",  1, enc); break;
-            case CYR_CAP_R:  rb_enc_str_buf_cat(dest, "R",  1, enc); break;
-            case CYR_CAP_S:  rb_enc_str_buf_cat(dest, "S",  1, enc); break;
-            case CYR_CAP_T:  rb_enc_str_buf_cat(dest, "T",  1, enc); break;
-            case CYR_CAP_U:  rb_enc_str_buf_cat(dest, "U",  1, enc); break;
-            case CYR_CAP_F:  rb_enc_str_buf_cat(dest, "F",  1, enc); break;
-            case CYR_CAP_H:  rb_enc_str_buf_cat(dest, "H",  1, enc); break;
-            case CYR_CAP_C:  rb_enc_str_buf_cat(dest, "C",  1, enc); break;
-            case CYR_CAP_TJ: STR_CAT_COND_ASCII(ascii, dest, "C", LAT_CAP_TJ, 1, enc); break;
-            case CYR_CAP_CH: STR_CAT_COND_ASCII(ascii, dest, "C", LAT_CAP_CH, 1, enc); break;
-            case CYR_CAP_ZH: STR_CAT_COND_ASCII(ascii, dest, "Z", LAT_CAP_ZH, 1, enc); break;
-            case CYR_CAP_SH: STR_CAT_COND_ASCII(ascii, dest, "S", LAT_CAP_SH, 1, enc); break;
-            case CYR_CAP_LJ:
-                rb_enc_str_buf_cat(dest, (force_upper ? "LJ" : "Lj"), 2, enc);
-                break;
-            case CYR_CAP_NJ:
-                rb_enc_str_buf_cat(dest, (force_upper ? "NJ" : "Nj"), 2, enc);
-                break;
-            case CYR_CAP_DJ:
-                STR_CAT_COND_ASCII(ascii, dest, (force_upper ? "DJ" : "Dj"), LAT_CAP_DJ, 2, enc);
-                break;
-            case CYR_CAP_DZ:
-                rb_enc_str_buf_cat(dest, "D", 1, enc);
-                if (force_upper) {
-                    STR_CAT_COND_ASCII(ascii, dest, "Z", LAT_CAP_ZH, 1, enc);
-                }
-                else {
-                    STR_CAT_COND_ASCII(ascii, dest, "z", LAT_ZH, 1, enc);
-                }
-                break;
-            default:
-                rb_enc_str_buf_cat(dest, pos, len, enc);
-            }
+        /* for cyrillic output, skip the second half of an input digraph */
+        if (strategy == 0 && is_digraph(out)) {
+            pos += next_len;
+            if (pos + len < end)
+                in2 = rb_enc_codepoint_len(pos + len, end, &next_len, enc);
         }
+
+        seen_cap = is_cap(in);
+
         pos += len;
         len = next_len;
-        codepoint = next_codepoint;
-        next_codepoint = 0;
+        in = in2;
     }
 
+    /* flush final sequence */
+    if (seq_start) rb_str_buf_cat(dest, seq_start, pos - seq_start);
+
     if (bang) {
         rb_str_shared_replace(str, dest);
     }
     else {
-	OBJ_INFECT(dest, str);
-	str = dest;
+        str = dest;
     }
 
     return str;
 }
 
+/**
+ * Returns a copy of <i>str</i> with Latin characters transliterated
+ * into Serbian Cyrillic.
+ *
+ * @overload to_cyrillic(str)
+ *   @param  [String] str text to be transliterated
+ *   @return [String] transliterated text
+ */
+static VALUE
+rb_str_to_cyrillic(VALUE self, VALUE str)
+{
+    return str_to_srb(str, 0, 0);
+}
+
+/**
+ * Performs transliteration of <code>Byk.to_cyrillic</code> in place,
+ * returning <i>str</i>, whether any changes were made or not.
+ *
+ * @overload to_cyrillic!(str)
+ *   @param  [String] str text to be transliterated
+ *   @return [String] transliterated text
+ */
 static VALUE
-rb_str_to_latin(int argc, VALUE *argv, VALUE str) {
-    return str_to_latin(argc, argv, str, 0, 0);
+rb_str_to_cyrillic_bang(VALUE self, VALUE str)
+{
+    return str_to_srb(str, 0, 1);
 }
 
+/**
+ * Returns a copy of <i>str</i> with Serbian Cyrillic characters
+ * transliterated into Latin.
+ *
+ * @overload to_latin(str)
+ *   @param  [String] str text to be transliterated
+ *   @return [String] transliterated text
+ */
 static VALUE
-rb_str_to_latin_bang(int argc, VALUE *argv, VALUE str) {
-    return str_to_latin(argc, argv, str, 0, 1);
+rb_str_to_latin(VALUE self, VALUE str)
+{
+    return str_to_srb(str, 1, 0);
 }
 
+/**
+ * Performs transliteration of <code>Byk.to_latin</code> in place,
+ * returning <i>str</i>, whether any changes were made or not.
+ *
+ * @overload to_latin!(str)
+ *   @param  [String] str text to be transliterated
+ *   @return [String] transliterated text
+ */
 static VALUE
-rb_str_to_ascii_latin(int argc, VALUE *argv, VALUE str) {
-    return str_to_latin(argc, argv, str, 1, 0);
+rb_str_to_latin_bang(VALUE self, VALUE str)
+{
+    return str_to_srb(str, 1, 1);
 }
 
+/**
+ * Returns a copy of <i>str</i> with Serbian characters transliterated
+ * into ASCII Latin.
+ *
+ * @overload to_ascii_latin(str)
+ *   @param  [String] str text to be transliterated
+ *   @return [String] transliterated text
+ */
 static VALUE
-rb_str_to_ascii_latin_bang(int argc, VALUE *argv, VALUE str) {
-    return str_to_latin(argc, argv, str, 1, 1);
+rb_str_to_ascii_latin(VALUE self, VALUE str)
+{
+    return str_to_srb(str, 2, 0);
+}
+
+/**
+ * Performs transliteration of <code>Byk.to_ascii_latin</code> in
+ * place, returning <i>str</i>, whether any changes were made or not.
+ *
+ * @overload to_ascii_latin!(str)
+ *   @param  [String] str text to be transliterated
+ *   @return [String] transliterated text
+ */
+static VALUE
+rb_str_to_ascii_latin_bang(VALUE self, VALUE str)
+{
+    return str_to_srb(str, 2, 1);
 }
 
 void Init_byk_native(void)
 {
-    rb_define_method(rb_cString, "to_latin",  rb_str_to_latin, -1);
-    rb_define_method(rb_cString, "to_latin!", rb_str_to_latin_bang, -1);
-    rb_define_method(rb_cString, "to_ascii_latin", rb_str_to_ascii_latin, -1);
-    rb_define_method(rb_cString, "to_ascii_latin!", rb_str_to_ascii_latin_bang, -1);
+    VALUE Byk = rb_define_module("Byk");
+    rb_define_singleton_method(Byk, "to_cyrillic", rb_str_to_cyrillic, 1);
+    rb_define_singleton_method(Byk, "to_cyrillic!", rb_str_to_cyrillic_bang, 1);
+    rb_define_singleton_method(Byk, "to_latin", rb_str_to_latin, 1);
+    rb_define_singleton_method(Byk, "to_latin!", rb_str_to_latin_bang, 1);
+    rb_define_singleton_method(Byk, "to_ascii_latin", rb_str_to_ascii_latin, 1);
+    rb_define_singleton_method(Byk, "to_ascii_latin!", rb_str_to_ascii_latin_bang, 1);
 }

data/lib/byk/core_ext/string.rb

@@ -0,0 +1,8 @@
+class String
+
+  Byk.singleton_methods.each do |method|
+    define_method(method) do
+      Byk.send(method, self)
+    end
+  end
+end

data/lib/byk/safe.rb

@@ -0,0 +1,14 @@
+# coding: utf-8
+
+require "byk_native"
+require "byk/version"
+
+module Byk
+
+  AZBUKA      = %w[а б в г д ђ е ж з и ј к л љ м н њ о п р с т ћ у ф х ц ч џ ш]
+  AZBUKA_CAPS = %W[А Б В Г Д Ђ Е Ж З И Ј К Л Љ М Н Њ О П Р С Т Ћ У Ф Х Ц Ч Џ Ш]
+
+  ABECEDA      = %w[a b c č ć d dž đ e f g h i j k l lj m n nj o p r s š t u v z ž]
+  ABECEDA_CAPS = %W[A B C Č Ć D Dž Đ E F G H I J K L Lj M N Nj O P R S Š T U V Z Ž]
+
+end

data/lib/byk/version.rb

@@ -1,3 +1,3 @@
 module Byk
-  VERSION = "0.4.0"
+  VERSION = "1.1.0"
 end

data/lib/byk.rb

@@ -1,14 +1,2 @@
-# coding: utf-8
-
-require "byk_native"
-require "byk/version"
-
-module Byk
-
-  AZBUKA      = %w[а б в г д ђ е ж з и ј к л љ м н њ о п р с т ћ у ф х ц ч џ ш]
-  AZBUKA_CAPS = %W[А Б В Г Д Ђ Е Ж З И Ј К Л Љ М Н Њ О П Р С Т Ћ У Ф Х Ц Ч Џ Ш]
-
-  ABECEDA      = %w[a b c č ć d dž đ e f g h i j k l lj m n nj o p r s š t u v z ž]
-  ABECEDA_CAPS = %W[A B C Č Ć D Dž Đ E F G H I J K L Lj M N Nj O P R S Š T U V Z Ž]
-
-end
+require "byk/safe"
+require "byk/core_ext/string"