--- /dev/null
+/*
+ * ISO-2022-JP <=> Unicode translate functions.
+ * by Norihisa Washitake <nori@washitake.com>
+ * US-ASCII/JIS X 0201/JIS X 0212 support
+ * by Hatuka*nezumi - IKEDA Soji <nezumi@jca.apc.org>
+ *
+ * $Id: iso2022jp.c,v 1.12 2004/05/23 14:28:24 mrsam Exp $
+ *
+ * This conversion is highly expensive, so it is recommended
+ * that you do not include iso-2022-jp supprt unless you need it.
+ */
+
+/*
+ * Debug Option.
+ * if you want to make iso2022jp test application,
+ * please set the value of _DEBUG to non-zero.
+ */
+#define JIS_DEBUG 0
+/* #define JIS_BUILD_APP */
+
+#include "iso2022jp.h"
+
+#if (JIS_DEBUG) > 0
+#ifdef JIS_BUILD_APP
+ #define JIS_OUT fprintf
+ #define JIS_OUT_FH stderr
+#else
+ #include <syslog.h>
+ #define JIS_OUT syslog
+ #define JIS_OUT_FH (LOG_MAIL|LOG_DEBUG)
+#endif
+#endif
+
+
+/*
+ * read_jis_char.
+ * -- from my second kanji conversion library in 2001. --
+ * Arguments:
+ * src: text in iso-2022-jp.
+ * ch: character info of each character.
+ * Returns:
+ * characters to be skipped in original text.
+ * this value is at least 1.
+ */
+
+static size_t read_jis_char(const char* src, struct jischar_t *ch)
+{
+ /*
+ * In most cases, JIS characters are grouped in 0x20
+ * characters. So we switch by value of src[0]/0x20.
+ */
+ switch (src[0] >> 5) {
+ case 0: /* 0x00 to 0x1F */
+ switch (src[0]) {
+ case JIS_CHAR_SI:
+ ch->type = JIS_TYPE_8BITKANA;
+ ch->value = 0;
+ return 1;
+ case JIS_CHAR_SO:
+ ch->type = JIS_TYPE_ASCII;
+ ch->value = 0;
+ return 1;
+ case JIS_CHAR_ESC:
+ ch->value = 0;
+ switch (src[1]) {
+ case '(': /* 94 character set (G0) */
+ switch (src[2]) {
+ case 'B': /* US-ASCII */
+ ch->type = JIS_TYPE_ASCII;
+ return 3;
+ case 'I': /* JIS X 0201 GR */
+ ch->type = JIS_TYPE_7BITKANA;
+ return 3;
+ case 'J': /* JIS X 0201 GL */
+ ch->type = JIS_TYPE_ROMAN;
+ return 3;
+ default:
+ ch->type = JIS_TYPE_ASCII;
+ ch->value = JIS_CHAR_ESC;
+ return 1;
+ }
+ case '$': /* 94/96n character set */
+ switch (src[2]) {
+ case '@': /* JIS C 6226:1978 */
+ ch->type = JIS_TYPE_JISX0208_1978;
+ return 3;
+ case 'B': /* JIS X 0208:1983/1990/1997 */
+ ch->type = JIS_TYPE_JISX0208;
+ return 3;
+ case '(':
+ switch (src[3]) {
+ case '@': /* JIS C 6226:1978 */
+ ch->type = JIS_TYPE_JISX0208_1978;
+ return 4;
+ case 'B': /* JIS X 0208:1983/1990/1997 */
+ ch->type = JIS_TYPE_JISX0208;
+ return 4;
+ case 'D': /* JIS X 0212:1990 */
+ ch->type = JIS_TYPE_JISX0212;
+ return 4;
+ default:
+ ch->type = JIS_TYPE_BINARY;
+ ch->value = JIS_CHAR_ESC;
+ return 1;
+ }
+ default:
+ ch->type = JIS_TYPE_BINARY;
+ ch->value = JIS_CHAR_ESC;
+ return 1;
+ }
+ case 'K': /* NEC KANJI (IN) */
+ ch->type = JIS_TYPE_JISX0208_1978;
+ return 1;
+ case 'H': /* NEC KANJI (OUT) */
+ ch->type = JIS_TYPE_ASCII;
+ return 1;
+ }
+ default:
+ ch->type = JIS_TYPE_BINARY;
+ ch->value = src[0];
+ return 1;
+ }
+ case 1: /* 0x20 to 0x3F */
+ case 2: /* 0x40 to 0x5F */
+ if (ch->type == JIS_TYPE_7BITKANA) {
+ ch->value = src[0] + 0x80;
+ return 1;
+ }
+ /* Other than 7bit kana are passed to next */
+ case 3: /* 0x60 to 0x7F */
+ if (src[0] == 0x7F) {
+ ch->type = JIS_TYPE_BINARY;
+ ch->value = src[0];
+ return 1;
+ }
+ if ((ch->type == JIS_TYPE_JISX0208
+ || ch->type == JIS_TYPE_JISX0208_1978
+ || ch->type == JIS_TYPE_JISX0212) && src[1]) {
+ ch->value = (src[0] * 0x100) + src[1];
+ return 2;
+ }
+ ch->value = src[0];
+ return 1;
+ case 4: /* 0x80 to 0x9F */
+ ch->value = src[0];
+ ch->type = JIS_TYPE_BINARY;
+ return 1;
+ case 5: /* 0xA0 to 0xBF */
+ case 6: /* 0xC0 to 0xDF */
+ if (ch->type == JIS_TYPE_8BITKANA) {
+ if (0xA0 < (unsigned)src[0] && (unsigned)src[0] <= 0xDF) {
+ ch->value = (unsigned char)src[0];
+ return 1;
+ }
+ }
+ ch->type = JIS_TYPE_BINARY;
+ ch->value = (unsigned char)src[0];
+ return 1;
+ case 7: /* 0xE0 to 0xFF */
+ ch->value = (unsigned char)src[0];
+ ch->type = JIS_TYPE_BINARY;
+ return 1;
+ default:
+ ch->value = (unsigned char)src[0];
+ ch->type = JIS_TYPE_BINARY;
+ return 1;
+ }
+}
+
+static unicode_char c2u_conv(int j, int jis_type)
+{
+ unsigned int upper = (j >> 8);
+ unsigned int lower = j & 0xFF;
+ const unicode_char **tbls;
+
+ if (!upper)
+ {
+ switch (jis_type)
+ {
+ /* JIS X 0201 GR */
+ case JIS_TYPE_7BITKANA:
+ case JIS_TYPE_8BITKANA:
+ if (0xA1 <= lower && lower <=0xDF)
+ return (unicode_char)(lower + (0xFF9F - 0xDF));
+ else
+ return (unicode_char)0xFFFD;
+ break;
+
+ /* JIS X 0201 GL */
+ case JIS_TYPE_ROMAN:
+ /* 2 characters replaced by JIS X 0201 */
+ if (lower == 0x5C) /* REVERSE SOLIDUS -> YEN SIGN */
+ return (unicode_char)0x00A5;
+ if (lower == 0x7E) /* TILDE -> OVERLINE */
+ return (unicode_char)0x203E;
+ /* break; */
+ /* US-ASCII or Control characters */
+ case JIS_TYPE_ASCII:
+ case JIS_TYPE_BINARY:
+ if (lower < 0x80)
+ return (unicode_char)lower;
+ else
+ return (unicode_char)0xFFFD;
+ break;
+
+ /* Otherwise return REPLACEMENT CHARACTER. */
+ default:
+ return (unicode_char)0xFFFD;
+ }
+ }
+
+ switch (jis_type)
+ {
+ /* JIS X 0208:1983/1990/1997 */
+ case JIS_TYPE_JISX0208:
+ tbls = jisx0208_to_uni_tbls;
+ break;
+
+ /* JIS C 6226:1978 */
+ case JIS_TYPE_JISX0208_1978:
+ tbls = jisx0208_1978_to_uni_tbls;
+ break;
+
+ /* JIS X 0212:1990 */
+ case JIS_TYPE_JISX0212:
+ tbls = jisx0212_to_uni_tbls;
+ break;
+
+ /* Otherwise return REPLACEMENT CHARACTER. */
+ default:
+ return (unicode_char)0xFFFD;
+ break;
+ }
+
+ if (0x20 < upper && upper < 0x7F
+ && 0x20 < lower && lower < 0x7F)
+ {
+ if (tbls[upper-0x21] != NULL
+ && tbls[upper-0x21][lower-0x21] != (unicode_char)0x003F)
+ {
+ if (tbls[upper-0x21][lower-0x21])
+ return tbls[upper-0x21][lower-0x21];
+ return (unicode_char)0xFFFD;
+ }
+ }
+
+ /* we should think of 8bit-JIS, maybe. */
+ /* but currently returns the REPLACEMENT CHARACTER. */
+ return (unicode_char)0xFFFD;
+}
+
+static unicode_char *c2u(const struct unicode_info *u,
+ const char *jis_str, int *err)
+{
+ size_t i, cnt, w;
+ unicode_char *uc;
+ struct jischar_t jchar;
+
+ if (err)
+ *err = -1;
+
+ /* Count the number of potential unicode characters first. */
+ i = cnt = 0;
+ jchar.type = 0;
+ jchar.value = 0;
+ while (jis_str[i]) {
+ i += read_jis_char(&jis_str[i], &jchar);
+ if (jchar.value)
+ ++cnt;
+ }
+
+ uc = malloc((cnt+1) * sizeof(unicode_char));
+#if JIS_DEBUG>0
+ if (uc)
+ JIS_OUT(JIS_OUT_FH, "c2u: allocated heap; 0x%04X bytes.\n", cnt+1);
+ else
+ JIS_OUT(JIS_OUT_FH, "c2u: heap allocation failed; 0x%04X bytes.\n", cnt+1);
+#endif
+ if (!uc)
+ return (NULL);
+
+ i = cnt = 0;
+ jchar.type = 0;
+ jchar.value = 0;
+ while (jis_str[i]) {
+ w = read_jis_char(&jis_str[i], &jchar);
+ if (jchar.value) {
+ uc[cnt] = c2u_conv(jchar.value, jchar.type);
+#if JIS_DEBUG > 1
+ JIS_OUT(JIS_OUT_FH, "c2u: converted; JIS 0x%04X => U+%04X", jchar.value, uc[cnt]);
+#endif
+ if (uc[cnt] == (unicode_char)0xFFFD && err)
+ {
+ *err = i;
+ free(uc);
+ return NULL;
+ }
+ ++cnt;
+ }
+ i+=w;
+ }
+
+ uc[cnt] = 0;
+#if JIS_DEBUG > 0
+ JIS_OUT(JIS_OUT_FH, "c2u: end of heap; 0x%04X bytes.", cnt+1);
+#endif
+ return (uc);
+}
+
+static void revlookup(unicode_char u, struct jischar_t *ch)
+{
+ unsigned int upper = u >> 8;
+ unsigned int lower = u & 0xff;
+
+ /* ISO-2022-JP(-1) is mapped inside BMP range. */
+ if (u >= (unicode_char)0x10000)
+ {
+ ch->type = JIS_TYPE_BINARY;
+ ch->value = 0x003F;
+ return;
+ }
+
+ /* US-ASCII */
+ if (u < (unicode_char)0x0080)
+ {
+ ch->type = JIS_TYPE_ASCII;
+ ch->value = (unsigned)u;
+ return;
+ }
+
+ /* 2 Characters replaced by JIS X 0201 */
+ if (u == (unicode_char)0x00a5)
+ {
+ ch->type = JIS_TYPE_ROMAN;
+ ch->value = 0x5C;
+ return;
+ }
+ if (u == (unicode_char)0x203E)
+ {
+ ch->type = JIS_TYPE_ROMAN;
+ ch->value = 0x7E;
+ return;
+ }
+
+ /* JIS X 0201 GR */
+ if ((unicode_char)0xFF61 <= u && u <= (unicode_char)0xFF9F)
+ {
+ ch->type = JIS_TYPE_8BITKANA;
+ ch->value = u - (unsigned)0xFF40 + (unsigned)0x80;
+ return;
+ }
+
+ /* JIS X 0208/JIS X 0212 */
+ if (uni_to_jisx0208_tbls[upper] != NULL
+ && uni_to_jisx0208_tbls[upper][lower] != 0x003F)
+ {
+ ch->type = JIS_TYPE_JISX0208;
+ ch->value = uni_to_jisx0208_tbls[upper][lower];
+ return;
+ }
+ if (uni_to_jisx0212_tbls[upper] != NULL
+ && uni_to_jisx0212_tbls[upper][lower] != 0x003F)
+ {
+ ch->type = JIS_TYPE_JISX0212;
+ ch->value = uni_to_jisx0212_tbls[upper][lower];
+ return;
+ }
+
+ /* return 'unknown' character if unknown */
+ ch->type = JIS_TYPE_BINARY;
+ ch->value = 0x003F;
+ return;
+}
+
+#if 0
+static int get_iso2022jp_type(unsigned j)
+{
+ if (0xA0 < j && j < 0xE0)
+ return JIS_TYPE_8BITKANA;
+ if (j > 0xff)
+ return JIS_TYPE_KANJI;
+ return JIS_TYPE_ASCII;
+}
+#endif
+
+static char *u2c(const struct unicode_info *u,
+ const unicode_char *str, int *err)
+{
+ size_t i, cnt;
+ int j;
+ int jtype = JIS_TYPE_ASCII;
+ int jt;
+ char *s;
+ struct jischar_t ch;
+
+ if (err)
+ *err = -1;
+
+ for (i = cnt = 0; str[i]; i++) {
+ revlookup(str[i], &ch);
+ jt = ch.type;
+ j = ch.value;
+ if (jt != jtype) {
+ cnt += ((jt == JIS_TYPE_JISX0212) ? 4 : 3);
+ jtype = jt;
+ }
+ cnt += ((jtype == JIS_TYPE_JISX0208 || jtype == JIS_TYPE_JISX0212) ? 2 : 1);
+ }
+ if (jtype != JIS_TYPE_ASCII && jtype != JIS_TYPE_BINARY)
+ cnt += 3;
+
+ s = malloc(cnt+1);
+#if JIS_DEBUG > 0
+ if (s)
+ JIS_OUT(JIS_OUT_FH, "u2c: allocated heap; 0x%04X bytes.\n", cnt+1);
+ else
+ JIS_OUT(JIS_OUT_FH, "u2c: heap allocation failed; 0x%04X bytes.\n", cnt+1);
+#endif
+ if (!s)
+ return (NULL);
+
+ jtype = JIS_TYPE_ASCII;
+ for (i = cnt = 0; str[i]; i++) {
+ revlookup(str[i], &ch);
+
+ jt = ch.type;
+ j = ch.value;
+ if (jt != jtype) {
+ switch (jt) {
+ case JIS_TYPE_JISX0208:
+ s[cnt++] = JIS_CHAR_ESC;
+ s[cnt++] = '$';
+ s[cnt++] = 'B';
+#if JIS_DEBUG > 2
+ JIS_OUT(JIS_OUT_FH, "u2c: changed map; JIS_TYPE_JISX0208.\n");
+#endif
+ break;
+ case JIS_TYPE_JISX0212:
+ s[cnt++] = JIS_CHAR_ESC;
+ s[cnt++] = '$';
+ s[cnt++] = '(';
+ s[cnt++] = 'D';
+ break;
+ case JIS_TYPE_7BITKANA:
+ case JIS_TYPE_8BITKANA:
+ s[cnt++] = JIS_CHAR_ESC;
+ s[cnt++] = '(';
+ s[cnt++] = 'I';
+#if JIS_DEBUG > 2
+ JIS_OUT(JIS_OUT_FH, "u2c: changed map; JIS_TYPE_8BITKANA.\n");
+#endif
+ break;
+ case JIS_TYPE_ROMAN:
+ s[cnt++] = JIS_CHAR_ESC;
+ s[cnt++] = '(';
+ s[cnt++] = 'J';
+ break;
+ default:
+ s[cnt++] = JIS_CHAR_ESC;
+ s[cnt++] = '(';
+ s[cnt++] = 'B';
+#if JIS_DEBUG > 2
+ JIS_OUT(JIS_OUT_FH, "u2c: changed map; JIS_TYPE_ASCII.\n");
+#endif
+ break;
+ }
+ jtype = jt;
+ }
+ switch (jtype) {
+ case JIS_TYPE_JISX0208:
+ case JIS_TYPE_JISX0212:
+ s[cnt++] = (char)(j >> 8);
+ s[cnt++] = (char)(j & 0xff);
+ break;
+ case JIS_TYPE_7BITKANA:
+ case JIS_TYPE_8BITKANA:
+ s[cnt++] = (char)(j - 0x80);
+ break;
+ default:
+ s[cnt++] = (char)j;
+ break;
+ }
+#if JIS_DEBUG > 1
+ JIS_OUT(JIS_OUT_FH, "u2c: converted; U+%04X => JIS 0x%04X\n", str[i], j);
+#endif
+ if (jtype == JIS_TYPE_BINARY && j == 0x003F)
+ if (err)
+ {
+ *err = i;
+ free(s);
+ return NULL;
+ }
+ }
+ if (jtype != JIS_TYPE_ASCII && jtype != JIS_TYPE_BINARY) {
+ s[cnt++] = JIS_CHAR_ESC;
+ s[cnt++] = '(';
+ s[cnt++] = 'B';
+ }
+ s[cnt] = '\x0';
+
+#if JIS_DEBUG > 0
+ JIS_OUT(JIS_OUT_FH, "u2c: end of heap; 0x%04X bytes.\n", cnt+1);
+#endif
+ return s;
+}
+
+static char *toupper_func(const struct unicode_info *u,
+ const char *cp, int *ip)
+{
+ unicode_char *uc = c2u(u, cp, ip);
+ char *s;
+ size_t i;
+
+ if (!uc)
+ return (NULL);
+
+ for (i=0; uc[i]; i++) {
+ if ((unicode_char)'a' <= uc[i] && uc[i] <= (unicode_char)'z')
+ uc[i] = uc[i] - ((unicode_char)'a' - (unicode_char)'A');
+ }
+
+ s = u2c(u, uc, NULL);
+ free(uc);
+ return (s);
+}
+
+static char *tolower_func(const struct unicode_info *u,
+ const char *cp, int *ip)
+{
+ unicode_char *uc = c2u(u, cp, ip);
+ char *s;
+ size_t i;
+
+ if (!uc)
+ return (NULL);
+
+ for (i=0; uc[i]; i++) {
+ if ((unicode_char)'A' <= uc[i] && uc[i] <= (unicode_char)'Z')
+ uc[i] = uc[i] + ((unicode_char)'a' - (unicode_char)'A');
+ }
+
+ s = u2c(u, uc, NULL);
+ free(uc);
+
+ return (s);
+}
+
+
+static char *totitle_func(const struct unicode_info *u,
+ const char *cp, int *ip)
+{
+ unicode_char *uc = c2u(u, cp, ip);
+ char *s;
+
+ if (!uc)
+ return (NULL);
+
+ /* Uh, sorry, what's "title" char? */
+ /*
+ * for (i=0; uc[i]; i++)
+ * uc[i] = unicode_tc(uc[i]);
+ */
+
+ s = u2c(u, uc, NULL);
+ free(uc);
+ return (s);
+}
+
+extern const struct unicode_info unicode_UTF8;
+
+const struct unicode_info unicode_ISO2022_JP = {
+ "ISO-2022-JP",
+ UNICODE_MB | UNICODE_REPLACEABLE | UNICODE_SISO |
+ UNICODE_HEADER_BASE64,
+ c2u,
+ u2c,
+ toupper_func,
+ tolower_func,
+ totitle_func,
+ &unicode_UTF8
+};
+
+const struct unicode_info unicode_ISO2022_JP_1 = {
+ "ISO-2022-JP-1",
+ UNICODE_MB | UNICODE_REPLACEABLE | UNICODE_SISO |
+ UNICODE_HEADER_BASE64,
+ c2u,
+ u2c,
+ toupper_func,
+ tolower_func,
+ totitle_func,
+ &unicode_UTF8
+};
+
+#if (JIS_DEBUG > 0) && defined(JIS_BUILD_APP)
+int main(int argc, char** argv)
+{
+ FILE* fp;
+ char c;
+ int cnt;
+ char* str;
+ unicode_char* ustr;
+ char* jstr;
+ int i;
+
+ if (argc<2) {
+ JIS_OUT(JIS_OUT_FH, "usage: %s filename(s)\n", argv[0]);
+ exit(1);
+ }
+
+ while (argc > 1) {
+ --argc;
+ JIS_OUT(JIS_OUT_FH, "main: opening file %s.\n", argv[argc]);
+ fp = fopen(argv[argc], "r");
+ cnt=0;
+ while (c = fgetc(fp) != EOF)
+ cnt++;
+
+ str = malloc(cnt+1);
+ fseek(fp, 0, SEEK_SET);
+ fread(str, cnt, 1, fp);
+ str[cnt] = 0;
+
+ ustr = c2u(str, NULL);
+ /* for (i=0; ustr[i]; i++)
+ * putchar(ustr[i]);
+ */
+ jstr = u2c(ustr, NULL);
+ for (i=0; jstr[i]; i++)
+ putchar(jstr[i]);
+
+ free(jstr);
+ free(ustr);
+ free(str);
+ }
+ return 1;
+}
+#endif /* defined(JIS_BUILD_APP) */