aboutsummaryrefslogtreecommitdiff
path: root/doc/libNX_X11/lcUniConv
diff options
context:
space:
mode:
authorMike Gabriel <mike.gabriel@das-netzwerkteam.de>2017-02-04 10:52:05 +0100
committerMihai Moldovan <ionic@ionic.de>2017-03-03 13:45:25 +0100
commit949f5e04efd80f2892f960f04a7454bc58f1d212 (patch)
tree0dd95b15b1f01eb89ad3ca284935a60c85aa5cf2 /doc/libNX_X11/lcUniConv
parentc9539417a0edf36f003378e398ad855fdeee5d0e (diff)
downloadnx-libs-949f5e04efd80f2892f960f04a7454bc58f1d212.tar.gz
nx-libs-949f5e04efd80f2892f960f04a7454bc58f1d212.tar.bz2
nx-libs-949f5e04efd80f2892f960f04a7454bc58f1d212.zip
doc/libNX_X11/lcUniConv: Move over the rather-documentary files 8bit_tab_to_h.c and cjk_tab_to_h.c to nx-libs's doc/ folder.
Diffstat (limited to 'doc/libNX_X11/lcUniConv')
-rw-r--r--doc/libNX_X11/lcUniConv/8bit_tab_to_h.c535
-rw-r--r--doc/libNX_X11/lcUniConv/cjk_tab_to_h.c1071
2 files changed, 1606 insertions, 0 deletions
diff --git a/doc/libNX_X11/lcUniConv/8bit_tab_to_h.c b/doc/libNX_X11/lcUniConv/8bit_tab_to_h.c
new file mode 100644
index 000000000..993979aeb
--- /dev/null
+++ b/doc/libNX_X11/lcUniConv/8bit_tab_to_h.c
@@ -0,0 +1,535 @@
+
+/*
+ * Generates an 8-bit character set table from a .TXT table as found on
+ * ftp.unicode.org or from a table containing the 256 Unicode values as
+ * hexadecimal integers.
+ * Examples:
+ *
+ * ./8bit_tab_to_h ISO-8859-1 iso8859_1 < tab8859_1
+ * ./8bit_tab_to_h ISO-8859-2 iso8859_2 < tab8859_2
+ * ./8bit_tab_to_h ISO-8859-3 iso8859_3 < tab8859_3
+ * ./8bit_tab_to_h ISO-8859-4 iso8859_4 < tab8859_4
+ * ./8bit_tab_to_h ISO-8859-5 iso8859_5 < tab8859_5
+ * ./8bit_tab_to_h ISO-8859-6 iso8859_6 < tab8859_6
+ * ./8bit_tab_to_h ISO-8859-7 iso8859_7 < tab8859_7
+ * ./8bit_tab_to_h ISO-8859-8 iso8859_8 < tab8859_8
+ * ./8bit_tab_to_h ISO-8859-9 iso8859_9 < tab8859_9
+ * ./8bit_tab_to_h ISO-8859-10 iso8859_10 < tab8859_10
+ * ./8bit_tab_to_h ISO-8859-14 iso8859_14 < tab8859_14
+ * ./8bit_tab_to_h ISO-8859-15 iso8859_15 < tab8859_15
+ * ./8bit_tab_to_h JISX0201.1976-0 jisx0201 < jis0201
+ * ./8bit_tab_to_h TIS620-0 tis620 < tabtis620
+ * ./8bit_tab_to_h KOI8-R koi8_r < tabkoi8_r
+ * ./8bit_tab_to_h KOI8-U koi8_u < tabkoi8_u
+ * ./8bit_tab_to_h ARMSCII-8 armscii_8 < tabarmscii_8
+ * ./8bit_tab_to_h CP1133 cp1133 < tabibm_cp1133
+ * ./8bit_tab_to_h MULELAO-1 mulelao < tabmulelao_1
+ * ./8bit_tab_to_h VISCII1.1-1 viscii1 < tabviscii
+ * ./8bit_tab_to_h TCVN-5712 tcvn < tabtcvn
+ * ./8bit_tab_to_h GEORGIAN-ACADEMY georgian_ac < tabgeorgian_academy
+ * ./8bit_tab_to_h GEORGIAN-PS georgian_ps < tabgeorgian_ps
+ *
+ * ./8bit_tab_to_h ISO-8859-1 iso8859_1 < 8859-1.TXT
+ * ./8bit_tab_to_h ISO-8859-2 iso8859_2 < 8859-2.TXT
+ * ./8bit_tab_to_h ISO-8859-3 iso8859_3 < 8859-3.TXT
+ * ./8bit_tab_to_h ISO-8859-4 iso8859_4 < 8859-4.TXT
+ * ./8bit_tab_to_h ISO-8859-5 iso8859_5 < 8859-5.TXT
+ * ./8bit_tab_to_h ISO-8859-6 iso8859_6 < 8859-6.TXT
+ * ./8bit_tab_to_h ISO-8859-7 iso8859_7 < 8859-7.TXT
+ * ./8bit_tab_to_h ISO-8859-8 iso8859_8 < 8859-8.TXT
+ * ./8bit_tab_to_h ISO-8859-9 iso8859_9 < 8859-9.TXT
+ * ./8bit_tab_to_h ISO-8859-10 iso8859_10 < 8859-10.TXT
+ * ./8bit_tab_to_h ISO-8859-14 iso8859_14 < 8859-14.TXT
+ * ./8bit_tab_to_h ISO-8859-15 iso8859_15 < 8859-15.TXT
+ * ./8bit_tab_to_h JISX0201.1976-0 jisx0201 < JIS0201.TXT
+ * ./8bit_tab_to_h KOI8-R koi8_r < KOI8-R.TXT
+ */
+
+#include <stdio.h>
+#include <stdlib.h>
+#include <stdbool.h>
+#include <string.h>
+
+int main (int argc, char *argv[])
+{
+ const char* charsetname;
+ const char* c_charsetname;
+ const char* filename;
+ const char* directory;
+ int charset2uni[0x100];
+
+ if (argc != 3 && argc != 4 && argc != 5)
+ exit(1);
+ charsetname = argv[1];
+ c_charsetname = argv[2];
+ if (argc > 3) {
+ filename = argv[3];
+ } else {
+ char* s = malloc(strlen(c_charsetname)+strlen(".h")+1);
+ strcpy(s,c_charsetname); strcat(s,".h");
+ filename = s;
+ }
+ directory = (argc > 4 ? argv[4] : "");
+
+ fprintf(stderr, "Creating %s%s\n", directory, filename);
+
+ {
+ int i, c;
+ c = getc(stdin);
+ ungetc(c,stdin);
+ if (c == '#') {
+ /* Read a unicode.org style .TXT file. */
+ for (i = 0; i < 0x100; i++)
+ charset2uni[i] = 0xfffd;
+ for (;;) {
+ c = getc(stdin);
+ if (c == EOF)
+ break;
+ if (c == '\n' || c == ' ' || c == '\t')
+ continue;
+ if (c == '#') {
+ do { c = getc(stdin); } while (!(c == EOF || c == '\n'));
+ continue;
+ }
+ ungetc(c,stdin);
+ if (scanf("0x%x", &i) != 1 || !(i >= 0 && i < 0x100))
+ exit(1);
+ do { c = getc(stdin); } while (c == ' ' || c == '\t');
+ if (c != EOF)
+ ungetc(c,stdin);
+ if (c == '\n' || c == '#')
+ continue;
+ if (scanf("0x%x", &charset2uni[i]) != 1)
+ exit(1);
+ }
+ } else {
+ /* Read a table of hexadecimal Unicode values. */
+ for (i = 0; i < 0x100; i++) {
+ if (scanf("%x", &charset2uni[i]) != 1)
+ exit(1);
+ if (charset2uni[i] < 0 || charset2uni[i] == 0xffff)
+ charset2uni[i] = 0xfffd;
+ }
+ if (scanf("%x", &i) != EOF)
+ exit(1);
+ }
+ }
+
+ /* Write the output file. */
+ {
+ FILE* f;
+
+ {
+ char* fname = malloc(strlen(directory)+strlen(filename)+1);
+ strcpy(fname,directory); strcat(fname,filename);
+ f = fopen(fname,"w");
+ if (f == NULL)
+ exit(1);
+ }
+
+ fprintf(f, "\n");
+ fprintf(f, "/*\n");
+ fprintf(f, " * %s\n", charsetname);
+ fprintf(f, " */\n");
+ fprintf(f, "\n");
+
+ {
+ int i, i1, i2, i3;
+ int line[16];
+ int tableno;
+ struct { int minline; int maxline; } tables[16];
+ bool some_invalid;
+ bool final_ret_reached;
+
+ for (i1 = 0; i1 < 16; i1++) {
+ bool all_invalid = true;
+ bool all_identity = true;
+ for (i2 = 0; i2 < 16; i2++) {
+ i = 16*i1+i2;
+ if (charset2uni[i] != 0xfffd)
+ all_invalid = false;
+ if (charset2uni[i] != i)
+ all_identity = false;
+ }
+ if (all_invalid)
+ line[i1] = -2;
+ else if (all_identity)
+ line[i1] = -1;
+ else
+ line[i1] = 0;
+ }
+ tableno = 0;
+ for (i1 = 0; i1 < 16; i1++) {
+ if (line[i1] >= 0) {
+ if (i1 > 0 && tableno > 0 && line[i1-1] == tableno-1) {
+ line[i1] = tableno-1;
+ tables[tableno-1].maxline = i1;
+ } else {
+ tableno++;
+ line[i1] = tableno-1;
+ tables[tableno-1].minline = tables[tableno-1].maxline = i1;
+ }
+ }
+ }
+ some_invalid = false;
+ for (i = 0; i < 0x100; i++)
+ if (charset2uni[i] == 0xfffd)
+ some_invalid = true;
+ if (tableno > 0) {
+ int t;
+ for (t = 0; t < tableno; t++) {
+ fprintf(f, "static const unsigned short %s_2uni", c_charsetname);
+ if (tableno > 1)
+ fprintf(f, "_%d", t+1);
+ fprintf(f, "[%d] = {\n", 16*(tables[t].maxline-tables[t].minline+1));
+ for (i1 = tables[t].minline; i1 <= tables[t].maxline; i1++) {
+ fprintf(f, " /* 0x%02x */\n", 16*i1);
+ for (i2 = 0; i2 < 2; i2++) {
+ fprintf(f, " ");
+ for (i3 = 0; i3 < 8; i3++) {
+ i = 16*i1+8*i2+i3;
+ fprintf(f, " 0x%04x,", charset2uni[i]);
+ }
+ fprintf(f, "\n");
+ }
+ }
+ fprintf(f, "};\n");
+ }
+ fprintf(f, "\n");
+ }
+ final_ret_reached = false;
+ fprintf(f, "static int\n%s_mbtowc (conv_t conv, ucs4_t *pwc, const unsigned char *s, int n)\n", c_charsetname);
+ fprintf(f, "{\n");
+ fprintf(f, " unsigned char c = *s;\n");
+ if (some_invalid) {
+ for (i1 = 0; i1 < 16;) {
+ int t = line[i1];
+ const char* indent;
+ for (i2 = i1; i2 < 16 && line[i2] == t; i2++);
+ indent = (i1 == 0 && i2 == 16 ? " " : " ");
+ if (i1 == 0) {
+ if (i2 == 16) {
+ } else {
+ fprintf(f, " if (c < 0x%02x) {\n", 16*i2);
+ }
+ } else {
+ if (i2 == 16) {
+ fprintf(f, " else {\n");
+ } else {
+ fprintf(f, " else if (c < 0x%02x) {\n", 16*i2);
+ }
+ }
+ if (t == -2) {
+ final_ret_reached = true;
+ } else if (t == -1) {
+ fprintf(f, "%s*pwc = (ucs4_t) c;\n", indent);
+ fprintf(f, "%sreturn 1;\n", indent);
+ } else {
+ fprintf(f, "%s", indent);
+ some_invalid = false;
+ for (i = 16*i1; i < 16*i2; i++)
+ if (charset2uni[i] == 0xfffd)
+ some_invalid = true;
+ if (some_invalid)
+ fprintf(f, "unsigned short wc = ");
+ else
+ fprintf(f, "*pwc = (ucs4_t) ");
+ fprintf(f, "%s_2uni", c_charsetname);
+ if (tableno > 1)
+ fprintf(f, "_%d", t+1);
+ fprintf(f, "[c");
+ if (tables[t].minline > 0)
+ fprintf(f, "-0x%02x", 16*tables[t].minline);
+ fprintf(f, "];\n");
+ if (some_invalid) {
+ fprintf(f, "%sif (wc != 0xfffd) {\n", indent);
+ fprintf(f, "%s *pwc = (ucs4_t) wc;\n", indent);
+ fprintf(f, "%s return 1;\n", indent);
+ fprintf(f, "%s}\n", indent);
+ final_ret_reached = true;
+ } else {
+ fprintf(f, "%sreturn 1;\n", indent);
+ }
+ }
+ if (!(i1 == 0 && i2 == 16))
+ fprintf(f, " }\n");
+ i1 = i2;
+ }
+ if (final_ret_reached)
+ fprintf(f, " return RET_ILSEQ;\n");
+ } else {
+ for (i1 = 0; i1 < 16;) {
+ int t = line[i1];
+ for (i2 = i1; i2 < 16 && line[i2] == t; i2++);
+ if (i1 == 0) {
+ if (i2 == 16) {
+ fprintf(f, " ");
+ } else {
+ fprintf(f, " if (c < 0x%02x)\n ", 16*i2);
+ }
+ } else {
+ if (i2 == 16) {
+ fprintf(f, " else\n ");
+ } else {
+ fprintf(f, " else if (c < 0x%02x)\n ", 16*i2);
+ }
+ }
+ if (t == -1)
+ fprintf(f, "*pwc = (ucs4_t) c;\n");
+ else {
+ fprintf(f, "*pwc = (ucs4_t) %s_2uni", c_charsetname);
+ if (tableno > 1)
+ fprintf(f, "_%d", t+1);
+ fprintf(f, "[c");
+ if (tables[t].minline > 0)
+ fprintf(f, "-0x%02x", 16*tables[t].minline);
+ fprintf(f, "];\n");
+ }
+ i1 = i2;
+ }
+ fprintf(f, " return 1;\n");
+ }
+ fprintf(f, "}\n");
+
+ }
+
+ fprintf(f, "\n");
+
+ {
+ int uni2charset[0x10000];
+ bool pages[0x100];
+ int line[0x2000];
+ int tableno;
+ struct { int minline; int maxline; int usecount; const char* suffix; } tables[0x2000];
+ bool need_c;
+ bool fix_0000;
+ int i, j, p, j1, j2, t;
+
+ for (j = 0; j < 0x10000; j++)
+ uni2charset[j] = 0;
+ for (p = 0; p < 0x100; p++)
+ pages[p] = false;
+ for (i = 0; i < 0x100; i++) {
+ j = charset2uni[i];
+ if (j != 0xfffd) {
+ uni2charset[j] = i;
+ pages[j>>8] = true;
+ }
+ }
+ for (j1 = 0; j1 < 0x2000; j1++) {
+ bool all_invalid = true;
+ bool all_identity = true;
+ for (j2 = 0; j2 < 8; j2++) {
+ j = 8*j1+j2;
+ if (uni2charset[j] != 0)
+ all_invalid = false;
+ if (uni2charset[j] != j)
+ all_identity = false;
+ }
+ if (all_invalid)
+ line[j1] = -2;
+ else if (all_identity)
+ line[j1] = -1;
+ else
+ line[j1] = 0;
+ }
+ tableno = 0;
+ for (j1 = 0; j1 < 0x2000; j1++) {
+ if (line[j1] >= 0) {
+ if (tableno > 0
+ && ((j1 > 0 && line[j1-1] == tableno-1)
+ || ((tables[tableno-1].maxline >> 5) == (j1 >> 5)
+ && j1 - tables[tableno-1].maxline <= 8))) {
+ line[j1] = tableno-1;
+ tables[tableno-1].maxline = j1;
+ } else {
+ tableno++;
+ line[j1] = tableno-1;
+ tables[tableno-1].minline = tables[tableno-1].maxline = j1;
+ }
+ }
+ }
+ for (t = 0; t < tableno; t++) {
+ tables[t].usecount = 0;
+ j1 = 8*tables[t].minline;
+ j2 = 8*(tables[t].maxline+1);
+ for (j = j1; j < j2; j++)
+ if (uni2charset[j] != 0)
+ tables[t].usecount++;
+ }
+ for (t = 0, p = -1, i = 0; t < tableno; t++) {
+ if (tables[t].usecount > 1) {
+ char* s;
+ if (p == tables[t].minline >> 5) {
+ s = malloc(5+1);
+ sprintf(s, "%02x_%d", p, ++i);
+ } else {
+ p = tables[t].minline >> 5;
+ s = malloc(2+1);
+ sprintf(s, "%02x", p);
+ }
+ tables[t].suffix = s;
+ } else
+ tables[t].suffix = NULL;
+ }
+ {
+ p = -1;
+ for (t = 0; t < tableno; t++)
+ if (tables[t].usecount > 1) {
+ p = 0;
+ fprintf(f, "static const unsigned char %s_page%s[%d] = {\n", c_charsetname, tables[t].suffix, 8*(tables[t].maxline-tables[t].minline+1));
+ for (j1 = tables[t].minline; j1 <= tables[t].maxline; j1++) {
+ if ((j1 % 0x20) == 0 && j1 > tables[t].minline)
+ fprintf(f, " /* 0x%04x */\n", 8*j1);
+ fprintf(f, " ");
+ for (j2 = 0; j2 < 8; j2++) {
+ j = 8*j1+j2;
+ fprintf(f, " 0x%02x,", uni2charset[j]);
+ }
+ fprintf(f, " /* 0x%02x-0x%02x */\n", 8*(j1 % 0x20), 8*(j1 % 0x20)+7);
+ }
+ fprintf(f, "};\n");
+ }
+ if (p >= 0)
+ fprintf(f, "\n");
+ }
+ need_c = false;
+ for (j1 = 0; j1 < 0x2000;) {
+ t = line[j1];
+ for (j2 = j1; j2 < 0x2000 && line[j2] == t; j2++);
+ if (t >= 0)
+ j2 = tables[t].maxline+1;
+ if (!(t == -2 || (t == -1 && j1 == 0)))
+ need_c = true;
+ j1 = j2;
+ }
+ fix_0000 = false;
+ fprintf(f, "static int\n%s_wctomb (conv_t conv, unsigned char *r, ucs4_t wc, int n)\n", c_charsetname);
+ fprintf(f, "{\n");
+ if (need_c)
+ fprintf(f, " unsigned char c = 0;\n");
+ for (j1 = 0; j1 < 0x2000;) {
+ t = line[j1];
+ for (j2 = j1; j2 < 0x2000 && line[j2] == t; j2++);
+ if (t >= 0) {
+ if (j1 != tables[t].minline) abort();
+ if (j2 > tables[t].maxline+1) abort();
+ j2 = tables[t].maxline+1;
+ }
+ if (t == -2) {
+ } else {
+ if (j1 == 0)
+ fprintf(f, " ");
+ else
+ fprintf(f, " else ");
+ if (t >= 0 && tables[t].usecount == 0) abort();
+ if (t >= 0 && tables[t].usecount == 1) {
+ if (j2 != j1+1) abort();
+ for (j = 8*j1; j < 8*j2; j++)
+ if (uni2charset[j] != 0) {
+ fprintf(f, "if (wc == 0x%04x)\n c = 0x%02x;\n", j, uni2charset[j]);
+ break;
+ }
+ } else {
+ if (j1 == 0) {
+ fprintf(f, "if (wc < 0x%04x)", 8*j2);
+ } else {
+ fprintf(f, "if (wc >= 0x%04x && wc < 0x%04x)", 8*j1, 8*j2);
+ }
+ if (t == -1) {
+ if (j1 == 0)
+ /* If wc == 0, the function must return 1, not -1. */
+ fprintf(f, " {\n *r = wc;\n return 1;\n }\n");
+ else
+ fprintf(f, "\n c = wc;\n");
+ } else {
+ fprintf(f, "\n c = %s_page%s[wc", c_charsetname, tables[t].suffix);
+ if (tables[t].minline > 0)
+ fprintf(f, "-0x%04x", 8*j1);
+ fprintf(f, "];\n");
+ if (j1 == 0 && uni2charset[0] == 0)
+ /* If wc == 0, the function must return 1, not -1. */
+ fix_0000 = true;
+ }
+ }
+ }
+ j1 = j2;
+ }
+ if (need_c) {
+ if (fix_0000)
+ fprintf(f, " if (c != 0 || wc == 0) {\n");
+ else
+ fprintf(f, " if (c != 0) {\n");
+ fprintf(f, " *r = c;\n");
+ fprintf(f, " return 1;\n");
+ fprintf(f, " }\n");
+ }
+ fprintf(f, " return RET_ILSEQ;\n");
+ fprintf(f, "}\n");
+
+ }
+
+ if (ferror(f) || fclose(f))
+ exit(1);
+ }
+
+#if 0
+
+ int i1, i2, i3, i1_min, i1_max, j1, j2;
+
+ i1_min = 16;
+ i1_max = -1;
+ for (i1 = 0; i1 < 16; i1++)
+ for (i2 = 0; i2 < 16; i2++)
+ if (charset2uni[16*i1+i2] != 0xfffd) {
+ if (i1_min > i1) i1_min = i1;
+ if (i1_max < i1) i1_max = i1;
+ }
+ printf("static const unsigned short %s_2uni[%d] = {\n",
+ name, 16*(i1_max-i1_min+1));
+ for (i1 = i1_min; i1 <= i1_max; i1++) {
+ printf(" /""* 0x%02x *""/\n", 16*i1);
+ for (i2 = 0; i2 < 2; i2++) {
+ printf(" ");
+ for (i3 = 0; i3 < 8; i3++) {
+ if (i3 > 0) printf(" ");
+ printf("0x%04x,", charset2uni[16*i1+8*i2+i3]);
+ }
+ printf("\n");
+ }
+ }
+ printf("};\n");
+ printf("\n");
+
+ for (p = 0; p < 0x100; p++)
+ pages[p] = 0;
+ for (i = 0; i < 0x100; i++)
+ if (charset2uni[i] != 0xfffd)
+ pages[charset2uni[i]>>8] = 1;
+ for (p = 0; p < 0x100; p++)
+ if (pages[p]) {
+ int j1_min = 32;
+ int j1_max = -1;
+ for (j1 = 0; j1 < 32; j1++)
+ for (j2 = 0; j2 < 8; j2++)
+ if (uni2charset[256*p+8*j1+j2] != 0) {
+ if (j1_min > j1) j1_min = j1;
+ if (j1_max < j1) j1_max = j1;
+ }
+ printf("static const unsigned char %s_page%02x[%d] = {\n",
+ name, p, 8*(j1_max-j1_min+1));
+ for (j1 = j1_min; j1 <= j1_max; j1++) {
+ printf(" ");
+ for (j2 = 0; j2 < 8; j2++)
+ printf("0x%02x, ", uni2charset[256*p+8*j1+j2]);
+ printf("/""* 0x%02x-0x%02x *""/\n", 8*j1, 8*j1+7);
+ }
+ printf("};\n");
+ }
+ printf("\n");
+
+}
+#endif
+
+ exit(0);
+}
diff --git a/doc/libNX_X11/lcUniConv/cjk_tab_to_h.c b/doc/libNX_X11/lcUniConv/cjk_tab_to_h.c
new file mode 100644
index 000000000..f70fe5e11
--- /dev/null
+++ b/doc/libNX_X11/lcUniConv/cjk_tab_to_h.c
@@ -0,0 +1,1071 @@
+
+/*
+ * Generates a CJK character set table from a .TXT table as found on
+ * ftp.unicode.org or in the X nls directory.
+ * Examples:
+ *
+ * ./cjk_tab_to_h GB2312.1980-0 gb2312 > gb2312.h < gb2312
+ * ./cjk_tab_to_h JISX0208.1983-0 jisx0208 > jisx0208.h < jis0208
+ * ./cjk_tab_to_h KSC5601.1987-0 ksc5601 > ksc5601.h < ksc5601
+ *
+ * ./cjk_tab_to_h GB2312.1980-0 gb2312 > gb2312.h < GB2312.TXT
+ * ./cjk_tab_to_h JISX0208.1983-0 jisx0208 > jisx0208.h < JIS0208.TXT
+ * ./cjk_tab_to_h JISX0212.1990-0 jisx0212 > jisx0212.h < JIS0212.TXT
+ * ./cjk_tab_to_h KSC5601.1987-0 ksc5601 > ksc5601.h < KSC5601.TXT
+ * ./cjk_tab_to_h KSX1001.1992-0 ksc5601 > ksc5601.h < KSX1001.TXT
+ *
+ * ./cjk_tab_to_h BIG5 big5 > big5.h < BIG5.TXT
+ *
+ * ./cjk_tab_to_h JOHAB johab > johab.h < JOHAB.TXT
+ *
+ * ./cjk_tab_to_h BIG5HKSCS-0 big5hkscs >big5hkscs.h < BIG5HKSCS.TXT
+ */
+
+#include <stdio.h>
+#include <stdlib.h>
+#include <stdbool.h>
+#include <string.h>
+
+typedef struct {
+ int start;
+ int end;
+} Block;
+
+typedef struct {
+ int rows; /* number of possible values for the 1st byte */
+ int cols; /* number of possible values for the 2nd byte */
+ int (*row_byte) (int row); /* returns the 1st byte value for a given row */
+ int (*col_byte) (int col); /* returns the 2nd byte value for a given col */
+ int (*byte_row) (int byte); /* converts a 1st byte value to a row, else -1 */
+ int (*byte_col) (int byte); /* converts a 2nd byte value to a col, else -1 */
+ const char* check_row_expr; /* format string for 1st byte value checking */
+ const char* check_col_expr; /* format string for 2nd byte value checking */
+ const char* byte_row_expr; /* format string for 1st byte value to row */
+ const char* byte_col_expr; /* format string for 2nd byte value to col */
+ int** charset2uni; /* charset2uni[0..rows-1][0..cols-1] is valid */
+ /* You'll understand the terms "row" and "col" when you buy Ken Lunde's book.
+ Once a row is fixed, choosing a "col" is the same as choosing a "cell". */
+ int* charsetpage; /* charsetpage[0..rows]: how large is a page for a row */
+ int ncharsetblocks;
+ Block* charsetblocks; /* blocks[0..nblocks-1] */
+ int* uni2charset; /* uni2charset[0x0000..0xffff] */
+} Encoding;
+
+/*
+ * Outputs the file title.
+ */
+static void output_title (const char *charsetname)
+{
+ printf("\n");
+ printf("/*\n");
+ printf(" * %s\n", charsetname);
+ printf(" */\n");
+ printf("\n");
+}
+
+/*
+ * Reads the charset2uni table from standard input.
+ */
+static void read_table (Encoding* enc)
+{
+ int row, col, i, i1, i2, c, j;
+
+ enc->charset2uni = malloc(enc->rows*sizeof(int*));
+ for (row = 0; row < enc->rows; row++)
+ enc->charset2uni[row] = malloc(enc->cols*sizeof(int));
+
+ for (row = 0; row < enc->rows; row++)
+ for (col = 0; col < enc->cols; col++)
+ enc->charset2uni[row][col] = 0xfffd;
+
+ c = getc(stdin);
+ ungetc(c,stdin);
+ if (c == '#') {
+ /* Read a unicode.org style .TXT file. */
+ for (;;) {
+ c = getc(stdin);
+ if (c == EOF)
+ break;
+ if (c == '\n' || c == ' ' || c == '\t')
+ continue;
+ if (c == '#') {
+ do { c = getc(stdin); } while (!(c == EOF || c == '\n'));
+ continue;
+ }
+ ungetc(c,stdin);
+ if (scanf("0x%x", &j) != 1)
+ exit(1);
+ i1 = j >> 8;
+ i2 = j & 0xff;
+ row = enc->byte_row(i1);
+ col = enc->byte_col(i2);
+ if (row < 0 || col < 0) {
+ fprintf(stderr, "lost entry for %02x %02x\n", i1, i2);
+ exit(1);
+ }
+ if (scanf(" 0x%x", &enc->charset2uni[row][col]) != 1)
+ exit(1);
+ }
+ } else {
+ /* Read a table of hexadecimal Unicode values. */
+ for (i1 = 32; i1 < 132; i1++)
+ for (i2 = 32; i2 < 132; i2++) {
+ i = scanf("%x", &j);
+ if (i == EOF)
+ goto read_done;
+ if (i != 1)
+ exit(1);
+ if (j < 0 || j == 0xffff)
+ j = 0xfffd;
+ if (j != 0xfffd) {
+ if (enc->byte_row(i1) < 0 || enc->byte_col(i2) < 0) {
+ fprintf(stderr, "lost entry at %02x %02x\n", i1, i2);
+ exit (1);
+ }
+ enc->charset2uni[enc->byte_row(i1)][enc->byte_col(i2)] = j;
+ }
+ }
+ read_done: ;
+ }
+}
+
+/*
+ * Computes the charsetpage[0..rows] array.
+ */
+static void find_charset2uni_pages (Encoding* enc)
+{
+ int row, col;
+
+ enc->charsetpage = malloc((enc->rows+1)*sizeof(int));
+
+ for (row = 0; row <= enc->rows; row++)
+ enc->charsetpage[row] = 0;
+
+ for (row = 0; row < enc->rows; row++) {
+ int used = 0;
+ for (col = 0; col < enc->cols; col++)
+ if (enc->charset2uni[row][col] != 0xfffd)
+ used = col+1;
+ enc->charsetpage[row] = used;
+ }
+}
+
+/*
+ * Fills in nblocks and blocks.
+ */
+static void find_charset2uni_blocks (Encoding* enc)
+{
+ int n, row, lastrow;
+
+ enc->charsetblocks = malloc(enc->rows*sizeof(Block));
+
+ n = 0;
+ for (row = 0; row < enc->rows; row++)
+ if (enc->charsetpage[row] > 0 && (row == 0 || enc->charsetpage[row-1] == 0)) {
+ for (lastrow = row; enc->charsetpage[lastrow+1] > 0; lastrow++);
+ enc->charsetblocks[n].start = row * enc->cols;
+ enc->charsetblocks[n].end = lastrow * enc->cols + enc->charsetpage[lastrow];
+ n++;
+ }
+ enc->ncharsetblocks = n;
+}
+
+/*
+ * Outputs the charset to unicode table and function.
+ */
+static void output_charset2uni (const char* name, Encoding* enc)
+{
+ int row, col, lastrow, col_max, i, i1_min, i1_max;
+
+ find_charset2uni_pages(enc);
+
+ find_charset2uni_blocks(enc);
+
+ for (row = 0; row < enc->rows; row++)
+ if (enc->charsetpage[row] > 0) {
+ if (row == 0 || enc->charsetpage[row-1] == 0) {
+ /* Start a new block. */
+ for (lastrow = row; enc->charsetpage[lastrow+1] > 0; lastrow++);
+ printf("static const unsigned short %s_2uni_page%02x[%d] = {\n",
+ name, enc->row_byte(row),
+ (lastrow-row) * enc->cols + enc->charsetpage[lastrow]);
+ }
+ printf(" /""* 0x%02x *""/\n ", enc->row_byte(row));
+ col_max = (enc->charsetpage[row+1] > 0 ? enc->cols : enc->charsetpage[row]);
+ for (col = 0; col < col_max; col++) {
+ printf(" 0x%04x,", enc->charset2uni[row][col]);
+ if ((col % 8) == 7 && (col+1 < col_max)) printf("\n ");
+ }
+ printf("\n");
+ if (enc->charsetpage[row+1] == 0) {
+ /* End a block. */
+ printf("};\n");
+ }
+ }
+ printf("\n");
+
+ printf("static int\n");
+ printf("%s_mbtowc (conv_t conv, ucs4_t *pwc, const unsigned char *s, int n)\n", name);
+ printf("{\n");
+ printf(" unsigned char c1 = s[0];\n");
+ printf(" if (");
+ for (i = 0; i < enc->ncharsetblocks; i++) {
+ i1_min = enc->row_byte(enc->charsetblocks[i].start / enc->cols);
+ i1_max = enc->row_byte((enc->charsetblocks[i].end-1) / enc->cols);
+ if (i > 0)
+ printf(" || ");
+ if (i1_min == i1_max)
+ printf("(c1 == 0x%02x)", i1_min);
+ else
+ printf("(c1 >= 0x%02x && c1 <= 0x%02x)", i1_min, i1_max);
+ }
+ printf(") {\n");
+ printf(" if (n >= 2) {\n");
+ printf(" unsigned char c2 = s[1];\n");
+ printf(" if (");
+ printf(enc->check_col_expr, "c2");
+ printf(") {\n");
+ printf(" unsigned int i = %d * (", enc->cols);
+ printf(enc->byte_row_expr, "c1");
+ printf(") + (");
+ printf(enc->byte_col_expr, "c2");
+ printf(");\n");
+ printf(" unsigned short wc = 0xfffd;\n");
+ for (i = 0; i < enc->ncharsetblocks; i++) {
+ printf(" ");
+ if (i > 0)
+ printf("} else ");
+ if (i < enc->ncharsetblocks-1)
+ printf("if (i < %d) ", enc->charsetblocks[i+1].start);
+ printf("{\n");
+ printf(" if (i < %d)\n", enc->charsetblocks[i].end);
+ printf(" wc = %s_2uni_page%02x[i", name, enc->row_byte(enc->charsetblocks[i].start / enc->cols));
+ if (enc->charsetblocks[i].start > 0)
+ printf("-%d", enc->charsetblocks[i].start);
+ printf("];\n");
+ }
+ printf(" }\n");
+ printf(" if (wc != 0xfffd) {\n");
+ printf(" *pwc = (ucs4_t) wc;\n");
+ printf(" return 2;\n");
+ printf(" }\n");
+ printf(" }\n");
+ printf(" return RET_ILSEQ;\n");
+ printf(" }\n");
+ printf(" return RET_TOOFEW(0);\n");
+ printf(" }\n");
+ printf(" return RET_ILSEQ;\n");
+ printf("}\n");
+ printf("\n");
+}
+
+/*
+ * Computes the uni2charset[0x0000..0xffff] array.
+ */
+static void invert (Encoding* enc)
+{
+ int row, col, j;
+
+ enc->uni2charset = malloc(0x10000*sizeof(int));
+
+ for (j = 0; j < 0x10000; j++)
+ enc->uni2charset[j] = 0;
+
+ for (row = 0; row < enc->rows; row++)
+ for (col = 0; col < enc->cols; col++) {
+ j = enc->charset2uni[row][col];
+ if (j != 0xfffd)
+ enc->uni2charset[j] = 0x100 * enc->row_byte(row) + enc->col_byte(col);
+ }
+}
+
+/*
+ * Outputs the unicode to charset table and function, using a linear array.
+ * (Suitable if the table is dense.)
+ */
+static void output_uni2charset_dense (const char* name, Encoding* enc)
+{
+ /* Like in 8bit_tab_to_h.c */
+ bool pages[0x100];
+ int line[0x2000];
+ int tableno;
+ struct { int minline; int maxline; int usecount; } tables[0x2000];
+ bool first;
+ int row, col, j, p, j1, j2, t;
+
+ for (p = 0; p < 0x100; p++)
+ pages[p] = false;
+ for (row = 0; row < enc->rows; row++)
+ for (col = 0; col < enc->cols; col++) {
+ j = enc->charset2uni[row][col];
+ if (j != 0xfffd)
+ pages[j>>8] = true;
+ }
+ for (j1 = 0; j1 < 0x2000; j1++) {
+ bool all_invalid = true;
+ for (j2 = 0; j2 < 8; j2++) {
+ j = 8*j1+j2;
+ if (enc->uni2charset[j] != 0)
+ all_invalid = false;
+ }
+ if (all_invalid)
+ line[j1] = -1;
+ else
+ line[j1] = 0;
+ }
+ tableno = 0;
+ for (j1 = 0; j1 < 0x2000; j1++) {
+ if (line[j1] >= 0) {
+ if (tableno > 0
+ && ((j1 > 0 && line[j1-1] == tableno-1)
+ || ((tables[tableno-1].maxline >> 5) == (j1 >> 5)
+ && j1 - tables[tableno-1].maxline <= 8))) {
+ line[j1] = tableno-1;
+ tables[tableno-1].maxline = j1;
+ } else {
+ tableno++;
+ line[j1] = tableno-1;
+ tables[tableno-1].minline = tables[tableno-1].maxline = j1;
+ }
+ }
+ }
+ for (t = 0; t < tableno; t++) {
+ tables[t].usecount = 0;
+ j1 = 8*tables[t].minline;
+ j2 = 8*(tables[t].maxline+1);
+ for (j = j1; j < j2; j++)
+ if (enc->uni2charset[j] != 0)
+ tables[t].usecount++;
+ }
+ {
+ p = -1;
+ for (t = 0; t < tableno; t++)
+ if (tables[t].usecount > 1) {
+ p = tables[t].minline >> 5;
+ printf("static const unsigned short %s_page%02x[%d] = {\n", name, p, 8*(tables[t].maxline-tables[t].minline+1));
+ for (j1 = tables[t].minline; j1 <= tables[t].maxline; j1++) {
+ if ((j1 % 0x20) == 0 && j1 > tables[t].minline)
+ printf(" /* 0x%04x */\n", 8*j1);
+ printf(" ");
+ for (j2 = 0; j2 < 8; j2++) {
+ j = 8*j1+j2;
+ printf(" 0x%04x,", enc->uni2charset[j]);
+ }
+ printf(" /*0x%02x-0x%02x*/\n", 8*(j1 % 0x20), 8*(j1 % 0x20)+7);
+ }
+ printf("};\n");
+ }
+ if (p >= 0)
+ printf("\n");
+ }
+ printf("static int\n%s_wctomb (conv_t conv, unsigned char *r, ucs4_t wc, int n)\n", name);
+ printf("{\n");
+ printf(" if (n >= 2) {\n");
+ printf(" unsigned short c = 0;\n");
+ first = true;
+ for (j1 = 0; j1 < 0x2000;) {
+ t = line[j1];
+ for (j2 = j1; j2 < 0x2000 && line[j2] == t; j2++);
+ if (t >= 0) {
+ if (j1 != tables[t].minline) abort();
+ if (j2 > tables[t].maxline+1) abort();
+ j2 = tables[t].maxline+1;
+ if (first)
+ printf(" ");
+ else
+ printf(" else ");
+ first = false;
+ if (tables[t].usecount == 0) abort();
+ if (tables[t].usecount == 1) {
+ if (j2 != j1+1) abort();
+ for (j = 8*j1; j < 8*j2; j++)
+ if (enc->uni2charset[j] != 0) {
+ printf("if (wc == 0x%04x)\n c = 0x%02x;\n", j, enc->uni2charset[j]);
+ break;
+ }
+ } else {
+ if (j1 == 0) {
+ printf("if (wc < 0x%04x)", 8*j2);
+ } else {
+ printf("if (wc >= 0x%04x && wc < 0x%04x)", 8*j1, 8*j2);
+ }
+ printf("\n c = %s_page%02x[wc", name, j1 >> 5);
+ if (tables[t].minline > 0)
+ printf("-0x%04x", 8*j1);
+ printf("];\n");
+ }
+ }
+ j1 = j2;
+ }
+ printf(" if (c != 0) {\n");
+ printf(" r[0] = (c >> 8); r[1] = (c & 0xff);\n");
+ printf(" return 2;\n");
+ printf(" }\n");
+ printf(" return RET_ILSEQ;\n");
+ printf(" }\n");
+ printf(" return RET_TOOSMALL;\n");
+ printf("}\n");
+}
+
+/*
+ * Outputs the unicode to charset table and function, using a packed array.
+ * (Suitable if the table is sparse.)
+ */
+static void output_uni2charset_sparse (const char* name, Encoding* enc)
+{
+ bool pages[0x100];
+ Block pageblocks[0x100]; int npageblocks;
+ int indx2charset[0x10000];
+ int summary_indx[0x1000];
+ int summary_used[0x1000];
+ int i, row, col, j, p, j1, j2, indx;
+
+ /* Fill pages[0x100]. */
+ for (p = 0; p < 0x100; p++)
+ pages[p] = false;
+ for (row = 0; row < enc->rows; row++)
+ for (col = 0; col < enc->cols; col++) {
+ j = enc->charset2uni[row][col];
+ if (j != 0xfffd)
+ pages[j>>8] = true;
+ }
+
+#if 0
+ for (p = 0; p < 0x100; p++)
+ if (pages[p]) {
+ printf("static const unsigned short %s_page%02x[256] = {\n", name, p);
+ for (j1 = 0; j1 < 32; j1++) {
+ printf(" ");
+ for (j2 = 0; j2 < 8; j2++)
+ printf("0x%04x, ", enc->uni2charset[256*p+8*j1+j2]);
+ printf("/""*0x%02x-0x%02x*""/\n", 8*j1, 8*j1+7);
+ }
+ printf("};\n");
+ }
+ printf("\n");
+#endif
+
+ /* Fill summary_indx[] and summary_used[]. */
+ indx = 0;
+ for (j1 = 0; j1 < 0x1000; j1++) {
+ summary_indx[j1] = indx;
+ summary_used[j1] = 0;
+ for (j2 = 0; j2 < 16; j2++) {
+ j = 16*j1+j2;
+ if (enc->uni2charset[j] != 0) {
+ indx2charset[indx++] = enc->uni2charset[j];
+ summary_used[j1] |= (1 << j2);
+ }
+ }
+ }
+
+ /* Fill npageblocks and pageblocks[]. */
+ npageblocks = 0;
+ for (p = 0; p < 0x100; ) {
+ if (pages[p] && (p == 0 || !pages[p-1])) {
+ pageblocks[npageblocks].start = 16*p;
+ do p++; while (p < 0x100 && pages[p]);
+ j1 = 16*p;
+ while (summary_used[j1-1] == 0) j1--;
+ pageblocks[npageblocks].end = j1;
+ npageblocks++;
+ } else
+ p++;
+ }
+
+ printf("static const unsigned short %s_2charset[%d] = {\n", name, indx);
+ for (i = 0; i < indx; ) {
+ if ((i % 8) == 0) printf(" ");
+ printf(" 0x%04x,", indx2charset[i]);
+ i++;
+ if ((i % 8) == 0 || i == indx) printf("\n");
+ }
+ printf("};\n");
+ printf("\n");
+ for (i = 0; i < npageblocks; i++) {
+ printf("static const Summary16 %s_uni2indx_page%02x[%d] = {\n", name,
+ pageblocks[i].start/16, pageblocks[i].end-pageblocks[i].start);
+ for (j1 = pageblocks[i].start; j1 < pageblocks[i].end; ) {
+ if (((16*j1) % 0x100) == 0) printf(" /""* 0x%04x *""/\n", 16*j1);
+ if ((j1 % 4) == 0) printf(" ");
+ printf(" { %4d, 0x%04x },", summary_indx[j1], summary_used[j1]);
+ j1++;
+ if ((j1 % 4) == 0 || j1 == pageblocks[i].end) printf("\n");
+ }
+ printf("};\n");
+ }
+ printf("\n");
+
+ printf("static int\n");
+ printf("%s_wctomb (conv_t conv, unsigned char *r, ucs4_t wc, int n)\n", name);
+ printf("{\n");
+ printf(" if (n >= 2) {\n");
+ printf(" const Summary16 *summary = NULL;\n");
+ for (i = 0; i < npageblocks; i++) {
+ printf(" ");
+ if (i > 0)
+ printf("else ");
+ printf("if (wc >= 0x%04x && wc < 0x%04x)\n",
+ 16*pageblocks[i].start, 16*pageblocks[i].end);
+ printf(" summary = &%s_uni2indx_page%02x[(wc>>4)", name,
+ pageblocks[i].start/16);
+ if (pageblocks[i].start > 0)
+ printf("-0x%03x", pageblocks[i].start);
+ printf("];\n");
+ }
+ printf(" if (summary) {\n");
+ printf(" unsigned short used = summary->used;\n");
+ printf(" unsigned int i = wc & 0x0f;\n");
+ printf(" if (used & ((unsigned short) 1 << i)) {\n");
+ printf(" unsigned short c;\n");
+ printf(" /* Keep in `used' only the bits 0..i-1. */\n");
+ printf(" used &= ((unsigned short) 1 << i) - 1;\n");
+ printf(" /* Add `summary->indx' and the number of bits set in `used'. */\n");
+ printf(" used = (used & 0x5555) + ((used & 0xaaaa) >> 1);\n");
+ printf(" used = (used & 0x3333) + ((used & 0xcccc) >> 2);\n");
+ printf(" used = (used & 0x0f0f) + ((used & 0xf0f0) >> 4);\n");
+ printf(" used = (used & 0x00ff) + (used >> 8);\n");
+ printf(" c = %s_2charset[summary->indx + used];\n", name);
+ printf(" r[0] = (c >> 8); r[1] = (c & 0xff);\n");
+ printf(" return 2;\n");
+ printf(" }\n");
+ printf(" }\n");
+ printf(" return RET_ILSEQ;\n");
+ printf(" }\n");
+ printf(" return RET_TOOSMALL;\n");
+ printf("}\n");
+}
+
+/* ISO-2022/EUC specifics */
+
+static int row_byte_normal (int row) { return 0x21+row; }
+static int col_byte_normal (int col) { return 0x21+col; }
+static int byte_row_normal (int byte) { return byte-0x21; }
+static int byte_col_normal (int byte) { return byte-0x21; }
+
+static void do_normal (const char* name)
+{
+ Encoding enc;
+
+ enc.rows = 94;
+ enc.cols = 94;
+ enc.row_byte = row_byte_normal;
+ enc.col_byte = col_byte_normal;
+ enc.byte_row = byte_row_normal;
+ enc.byte_col = byte_col_normal;
+ enc.check_row_expr = "%1$s >= 0x21 && %1$s < 0x7f";
+ enc.check_col_expr = "%1$s >= 0x21 && %1$s < 0x7f";
+ enc.byte_row_expr = "%1$s - 0x21";
+ enc.byte_col_expr = "%1$s - 0x21";
+
+ read_table(&enc);
+ output_charset2uni(name,&enc);
+ invert(&enc); output_uni2charset_sparse(name,&enc);
+}
+
+/* Note: On first sight, the jisx0212_2charset[] table seems to be in order,
+ starting from the charset=0x3021/uni=0x4e02 pair. But it's only mostly in
+ order. There are 75 out-of-order values, scattered all throughout the table.
+ */
+
+static void do_normal_only_charset2uni (const char* name)
+{
+ Encoding enc;
+
+ enc.rows = 94;
+ enc.cols = 94;
+ enc.row_byte = row_byte_normal;
+ enc.col_byte = col_byte_normal;
+ enc.byte_row = byte_row_normal;
+ enc.byte_col = byte_col_normal;
+ enc.check_row_expr = "%1$s >= 0x21 && %1$s < 0x7f";
+ enc.check_col_expr = "%1$s >= 0x21 && %1$s < 0x7f";
+ enc.byte_row_expr = "%1$s - 0x21";
+ enc.byte_col_expr = "%1$s - 0x21";
+
+ read_table(&enc);
+ output_charset2uni(name,&enc);
+}
+
+/* CNS 11643 specifics - trick to put two tables into one */
+
+static int row_byte_cns11643 (int row) {
+ return 0x100 * (row / 94) + (row % 94) + 0x21;
+}
+static int byte_row_cns11643 (int byte) {
+ return (byte >= 0x100 && byte < 0x200 ? byte-0x121 :
+ byte >= 0x200 && byte < 0x300 ? byte-0x221+94 :
+ byte >= 0x300 && byte < 0x400 ? byte-0x321+2*94 :
+ -1);
+}
+
+static void do_cns11643_only_uni2charset (const char* name)
+{
+ Encoding enc;
+ int j, x;
+
+ enc.rows = 3*94;
+ enc.cols = 94;
+ enc.row_byte = row_byte_cns11643;
+ enc.col_byte = col_byte_normal;
+ enc.byte_row = byte_row_cns11643;
+ enc.byte_col = byte_col_normal;
+ enc.check_row_expr = "%1$s >= 0x21 && %1$s < 0x7f";
+ enc.check_col_expr = "%1$s >= 0x21 && %1$s < 0x7f";
+ enc.byte_row_expr = "%1$s - 0x21";
+ enc.byte_col_expr = "%1$s - 0x21";
+
+ read_table(&enc);
+ invert(&enc);
+ /* Move the 2 plane bits into the unused bits 15 and 7. */
+ for (j = 0; j < 0x10000; j++) {
+ x = enc.uni2charset[j];
+ if (x != 0) {
+ if (x & 0x8080) abort();
+ switch (x >> 16) {
+ case 0: /* plane 1 */ x = (x & 0xffff) | 0x0000; break;
+ case 1: /* plane 2 */ x = (x & 0xffff) | 0x0080; break;
+ case 2: /* plane 3 */ x = (x & 0xffff) | 0x8000; break;
+ default: abort();
+ }
+ enc.uni2charset[j] = x;
+ }
+ }
+ output_uni2charset_sparse(name,&enc);
+}
+
+/* GBK specifics */
+
+static int row_byte_gbk1 (int row) {
+ return 0x81+row;
+}
+static int col_byte_gbk1 (int col) {
+ return (col >= 0x3f ? 0x41 : 0x40) + col;
+}
+static int byte_row_gbk1 (int byte) {
+ if (byte >= 0x81 && byte < 0xff)
+ return byte-0x81;
+ else
+ return -1;
+}
+static int byte_col_gbk1 (int byte) {
+ if (byte >= 0x40 && byte < 0x7f)
+ return byte-0x40;
+ else if (byte >= 0x80 && byte < 0xff)
+ return byte-0x41;
+ else
+ return -1;
+}
+
+static void do_gbk1 (const char* name)
+{
+ Encoding enc;
+
+ enc.rows = 126;
+ enc.cols = 190;
+ enc.row_byte = row_byte_gbk1;
+ enc.col_byte = col_byte_gbk1;
+ enc.byte_row = byte_row_gbk1;
+ enc.byte_col = byte_col_gbk1;
+ enc.check_row_expr = "%1$s >= 0x81 && %1$s < 0xff";
+ enc.check_col_expr = "(%1$s >= 0x40 && %1$s < 0x7f) || (%1$s >= 0x80 && %1$s < 0xff)";
+ enc.byte_row_expr = "%1$s - 0x81";
+ enc.byte_col_expr = "%1$s - (%1$s >= 0x80 ? 0x41 : 0x40)";
+
+ read_table(&enc);
+ output_charset2uni(name,&enc);
+ invert(&enc); output_uni2charset_dense(name,&enc);
+}
+
+static void do_gbk1_only_charset2uni (const char* name)
+{
+ Encoding enc;
+
+ enc.rows = 126;
+ enc.cols = 190;
+ enc.row_byte = row_byte_gbk1;
+ enc.col_byte = col_byte_gbk1;
+ enc.byte_row = byte_row_gbk1;
+ enc.byte_col = byte_col_gbk1;
+ enc.check_row_expr = "%1$s >= 0x81 && %1$s < 0xff";
+ enc.check_col_expr = "(%1$s >= 0x40 && %1$s < 0x7f) || (%1$s >= 0x80 && %1$s < 0xff)";
+ enc.byte_row_expr = "%1$s - 0x81";
+ enc.byte_col_expr = "%1$s - (%1$s >= 0x80 ? 0x41 : 0x40)";
+
+ read_table(&enc);
+ output_charset2uni(name,&enc);
+}
+
+static int row_byte_gbk2 (int row) {
+ return 0x81+row;
+}
+static int col_byte_gbk2 (int col) {
+ return (col >= 0x3f ? 0x41 : 0x40) + col;
+}
+static int byte_row_gbk2 (int byte) {
+ if (byte >= 0x81 && byte < 0xff)
+ return byte-0x81;
+ else
+ return -1;
+}
+static int byte_col_gbk2 (int byte) {
+ if (byte >= 0x40 && byte < 0x7f)
+ return byte-0x40;
+ else if (byte >= 0x80 && byte < 0xa1)
+ return byte-0x41;
+ else
+ return -1;
+}
+
+static void do_gbk2_only_charset2uni (const char* name)
+{
+ Encoding enc;
+
+ enc.rows = 126;
+ enc.cols = 96;
+ enc.row_byte = row_byte_gbk2;
+ enc.col_byte = col_byte_gbk2;
+ enc.byte_row = byte_row_gbk2;
+ enc.byte_col = byte_col_gbk2;
+ enc.check_row_expr = "%1$s >= 0x81 && %1$s < 0xff";
+ enc.check_col_expr = "(%1$s >= 0x40 && %1$s < 0x7f) || (%1$s >= 0x80 && %1$s < 0xa1)";
+ enc.byte_row_expr = "%1$s - 0x81";
+ enc.byte_col_expr = "%1$s - (%1$s >= 0x80 ? 0x41 : 0x40)";
+
+ read_table(&enc);
+ output_charset2uni(name,&enc);
+}
+
+static void do_gbk1_only_uni2charset (const char* name)
+{
+ Encoding enc;
+
+ enc.rows = 126;
+ enc.cols = 190;
+ enc.row_byte = row_byte_gbk1;
+ enc.col_byte = col_byte_gbk1;
+ enc.byte_row = byte_row_gbk1;
+ enc.byte_col = byte_col_gbk1;
+ enc.check_row_expr = "%1$s >= 0x81 && %1$s < 0xff";
+ enc.check_col_expr = "(%1$s >= 0x40 && %1$s < 0x7f) || (%1$s >= 0x80 && %1$s < 0xff)";
+ enc.byte_row_expr = "%1$s - 0x81";
+ enc.byte_col_expr = "%1$s - (%1$s >= 0x80 ? 0x41 : 0x40)";
+
+ read_table(&enc);
+ invert(&enc); output_uni2charset_sparse(name,&enc);
+}
+
+/* KSC 5601 specifics */
+
+/*
+ * Reads the charset2uni table from standard input.
+ */
+static void read_table_ksc5601 (Encoding* enc)
+{
+ int row, col, i, i1, i2, c, j;
+
+ enc->charset2uni = malloc(enc->rows*sizeof(int*));
+ for (row = 0; row < enc->rows; row++)
+ enc->charset2uni[row] = malloc(enc->cols*sizeof(int));
+
+ for (row = 0; row < enc->rows; row++)
+ for (col = 0; col < enc->cols; col++)
+ enc->charset2uni[row][col] = 0xfffd;
+
+ c = getc(stdin);
+ ungetc(c,stdin);
+ if (c == '#') {
+ /* Read a unicode.org style .TXT file. */
+ for (;;) {
+ c = getc(stdin);
+ if (c == EOF)
+ break;
+ if (c == '\n' || c == ' ' || c == '\t')
+ continue;
+ if (c == '#') {
+ do { c = getc(stdin); } while (!(c == EOF || c == '\n'));
+ continue;
+ }
+ ungetc(c,stdin);
+ if (scanf("0x%x", &j) != 1)
+ exit(1);
+ i1 = j >> 8;
+ i2 = j & 0xff;
+ if (scanf(" 0x%x", &j) != 1)
+ exit(1);
+ /* Take only the range covered by KS C 5601.1987-0 = KS C 5601.1989-0
+ = KS X 1001.1992, ignore the rest. */
+ if (!(i1 >= 128+33 && i1 < 128+127 && i2 >= 128+33 && i2 < 128+127))
+ continue; /* KSC5601 specific */
+ i1 &= 0x7f; /* KSC5601 specific */
+ i2 &= 0x7f; /* KSC5601 specific */
+ row = enc->byte_row(i1);
+ col = enc->byte_col(i2);
+ if (row < 0 || col < 0) {
+ fprintf(stderr, "lost entry for %02x %02x\n", i1, i2);
+ exit(1);
+ }
+ enc->charset2uni[row][col] = j;
+ }
+ } else {
+ /* Read a table of hexadecimal Unicode values. */
+ for (i1 = 33; i1 < 127; i1++)
+ for (i2 = 33; i2 < 127; i2++) {
+ i = scanf("%x", &j);
+ if (i == EOF)
+ goto read_done;
+ if (i != 1)
+ exit(1);
+ if (j < 0 || j == 0xffff)
+ j = 0xfffd;
+ if (j != 0xfffd) {
+ if (enc->byte_row(i1) < 0 || enc->byte_col(i2) < 0) {
+ fprintf(stderr, "lost entry at %02x %02x\n", i1, i2);
+ exit (1);
+ }
+ enc->charset2uni[enc->byte_row(i1)][enc->byte_col(i2)] = j;
+ }
+ }
+ read_done: ;
+ }
+}
+
+static void do_ksc5601 (const char* name)
+{
+ Encoding enc;
+
+ enc.rows = 94;
+ enc.cols = 94;
+ enc.row_byte = row_byte_normal;
+ enc.col_byte = col_byte_normal;
+ enc.byte_row = byte_row_normal;
+ enc.byte_col = byte_col_normal;
+ enc.check_row_expr = "%1$s >= 0x21 && %1$s < 0x7f";
+ enc.check_col_expr = "%1$s >= 0x21 && %1$s < 0x7f";
+ enc.byte_row_expr = "%1$s - 0x21";
+ enc.byte_col_expr = "%1$s - 0x21";
+
+ read_table_ksc5601(&enc);
+ output_charset2uni(name,&enc);
+ invert(&enc); output_uni2charset_sparse(name,&enc);
+}
+
+/* Big5 specifics */
+
+static int row_byte_big5 (int row) {
+ return 0xa1+row;
+}
+static int col_byte_big5 (int col) {
+ return (col >= 0x3f ? 0x62 : 0x40) + col;
+}
+static int byte_row_big5 (int byte) {
+ if (byte >= 0xa1 && byte < 0xff)
+ return byte-0xa1;
+ else
+ return -1;
+}
+static int byte_col_big5 (int byte) {
+ if (byte >= 0x40 && byte < 0x7f)
+ return byte-0x40;
+ else if (byte >= 0xa1 && byte < 0xff)
+ return byte-0x62;
+ else
+ return -1;
+}
+
+static void do_big5 (const char* name)
+{
+ Encoding enc;
+
+ enc.rows = 94;
+ enc.cols = 157;
+ enc.row_byte = row_byte_big5;
+ enc.col_byte = col_byte_big5;
+ enc.byte_row = byte_row_big5;
+ enc.byte_col = byte_col_big5;
+ enc.check_row_expr = "%1$s >= 0xa1 && %1$s < 0xff";
+ enc.check_col_expr = "(%1$s >= 0x40 && %1$s < 0x7f) || (%1$s >= 0xa1 && %1$s < 0xff)";
+ enc.byte_row_expr = "%1$s - 0xa1";
+ enc.byte_col_expr = "%1$s - (%1$s >= 0xa1 ? 0x62 : 0x40)";
+
+ read_table(&enc);
+ output_charset2uni(name,&enc);
+ invert(&enc); output_uni2charset_sparse(name,&enc);
+}
+
+/* Big5-HKSCS specifics */
+
+static int row_byte_big5hkscs (int row) {
+ return 0x81+row;
+}
+static int col_byte_big5hkscs (int col) {
+ return (col >= 0x3f ? 0x62 : 0x40) + col;
+}
+static int byte_row_big5hkscs (int byte) {
+ if (byte >= 0x81 && byte < 0xff)
+ return byte-0x81;
+ else
+ return -1;
+}
+static int byte_col_big5hkscs (int byte) {
+ if (byte >= 0x40 && byte < 0x7f)
+ return byte-0x40;
+ else if (byte >= 0xa1 && byte < 0xff)
+ return byte-0x62;
+ else
+ return -1;
+}
+
+static void do_big5hkscs (const char* name)
+{
+ Encoding enc;
+
+ enc.rows = 126;
+ enc.cols = 157;
+ enc.row_byte = row_byte_big5hkscs;
+ enc.col_byte = col_byte_big5hkscs;
+ enc.byte_row = byte_row_big5hkscs;
+ enc.byte_col = byte_col_big5hkscs;
+ enc.check_row_expr = "%1$s >= 0x81 && %1$s < 0xff";
+ enc.check_col_expr = "(%1$s >= 0x40 && %1$s < 0x7f) || (%1$s >= 0xa1 && %1$s < 0xff)";
+ enc.byte_row_expr = "%1$s - 0x81";
+ enc.byte_col_expr = "%1$s - (%1$s >= 0xa1 ? 0x62 : 0x40)";
+
+ read_table(&enc);
+ output_charset2uni(name,&enc);
+ invert(&enc); output_uni2charset_sparse(name,&enc);
+}
+
+/* Johab Hangul specifics */
+
+static int row_byte_johab_hangul (int row) {
+ return 0x84+row;
+}
+static int col_byte_johab_hangul (int col) {
+ return (col >= 0x3e ? 0x43 : 0x41) + col;
+}
+static int byte_row_johab_hangul (int byte) {
+ if (byte >= 0x84 && byte < 0xd4)
+ return byte-0x84;
+ else
+ return -1;
+}
+static int byte_col_johab_hangul (int byte) {
+ if (byte >= 0x41 && byte < 0x7f)
+ return byte-0x41;
+ else if (byte >= 0x81 && byte < 0xff)
+ return byte-0x43;
+ else
+ return -1;
+}
+
+static void do_johab_hangul (const char* name)
+{
+ Encoding enc;
+
+ enc.rows = 80;
+ enc.cols = 188;
+ enc.row_byte = row_byte_johab_hangul;
+ enc.col_byte = col_byte_johab_hangul;
+ enc.byte_row = byte_row_johab_hangul;
+ enc.byte_col = byte_col_johab_hangul;
+ enc.check_row_expr = "%1$s >= 0x84 && %1$s < 0xd4";
+ enc.check_col_expr = "(%1$s >= 0x41 && %1$s < 0x7f) || (%1$s >= 0x81 && %1$s < 0xff)";
+ enc.byte_row_expr = "%1$s - 0x84";
+ enc.byte_col_expr = "%1$s - (%1$s >= 0x81 ? 0x43 : 0x41)";
+
+ read_table(&enc);
+ output_charset2uni(name,&enc);
+ invert(&enc); output_uni2charset_dense(name,&enc);
+}
+
+/* SJIS specifics */
+
+static int row_byte_sjis (int row) {
+ return (row >= 0x1f ? 0xc1 : 0x81) + row;
+}
+static int col_byte_sjis (int col) {
+ return (col >= 0x3f ? 0x41 : 0x40) + col;
+}
+static int byte_row_sjis (int byte) {
+ if (byte >= 0x81 && byte < 0xa0)
+ return byte-0x81;
+ else if (byte >= 0xe0)
+ return byte-0xc1;
+ else
+ return -1;
+}
+static int byte_col_sjis (int byte) {
+ if (byte >= 0x40 && byte < 0x7f)
+ return byte-0x40;
+ else if (byte >= 0x80 && byte < 0xfd)
+ return byte-0x41;
+ else
+ return -1;
+}
+
+static void do_sjis (const char* name)
+{
+ Encoding enc;
+
+ enc.rows = 94;
+ enc.cols = 188;
+ enc.row_byte = row_byte_sjis;
+ enc.col_byte = col_byte_sjis;
+ enc.byte_row = byte_row_sjis;
+ enc.byte_col = byte_col_sjis;
+ enc.check_row_expr = "(%1$s >= 0x81 && %1$s < 0xa0) || (%1$s >= 0xe0)";
+ enc.check_col_expr = "(%1$s >= 0x40 && %1$s < 0x7f) || (%1$s >= 0x80 && %1$s < 0xfd)";
+ enc.byte_row_expr = "%1$s - (%1$s >= 0xe0 ? 0xc1 : 0x81)";
+ enc.byte_col_expr = "%1$s - (%1$s >= 0x80 ? 0x41 : 0x40)";
+
+ read_table(&enc);
+ output_charset2uni(name,&enc);
+ invert(&enc); output_uni2charset_sparse(name,&enc);
+}
+
+/* Main program */
+
+int main (int argc, char *argv[])
+{
+ const char* charsetname;
+ const char* name;
+
+ if (argc != 3)
+ exit(1);
+ charsetname = argv[1];
+ name = argv[2];
+
+ output_title(charsetname);
+
+ if (!strcmp(name,"gb2312") || !strcmp(name,"gb12345ext")
+ || !strcmp(name,"jisx0208") || !strcmp(name,"jisx0212"))
+ do_normal(name);
+ else if (!strcmp(name,"cns11643_1") || !strcmp(name,"cns11643_2")
+ || !strcmp(name,"cns11643_3"))
+ do_normal_only_charset2uni(name);
+ else if (!strcmp(name,"cns11643_inv"))
+ do_cns11643_only_uni2charset(name);
+ else if (!strcmp(name,"gbkext1"))
+ do_gbk1_only_charset2uni(name);
+ else if (!strcmp(name,"gbkext2"))
+ do_gbk2_only_charset2uni(name);
+ else if (!strcmp(name,"gbkext_inv"))
+ do_gbk1_only_uni2charset(name);
+ else if (!strcmp(name,"cp936ext"))
+ do_gbk1(name);
+ else if (!strcmp(name,"ksc5601"))
+ do_ksc5601(name);
+ else if (!strcmp(name,"big5") || !strcmp(name,"cp950ext"))
+ do_big5(name);
+ else if (!strcmp(name,"big5hkscs"))
+ do_big5hkscs(name);
+ else if (!strcmp(name,"johab_hangul"))
+ do_johab_hangul(name);
+ else if (!strcmp(name,"cp932ext"))
+ do_sjis(name);
+ else
+ exit(1);
+
+ return 0;
+}