mendexk: Add a new option -I to switch internal buffers to EUC-JP or UTF-8

git-svn-id: svn://tug.org/texlive/trunk@33962 c570f23f-e606-0410-a88d-b1316a301751

15 files changed, 214 insertions, 39 deletions

diff --git a/Build/source/texk/mendexk/ChangeLog b/Build/source/texk/mendexk/ChangeLog
index 63dabfbdbc2..46d0c29a4a5 100644
--- a/Build/source/texk/mendexk/ChangeLog
+++ b/Build/source/texk/mendexk/ChangeLog
@@ -1,3 +1,14 @@
+2014-05-11  TANAKA Takuji  <KXD02663@nifty.ne.jp>
+
+	* main.c: Add TeX Live version.
+	* {main,convert,fread,fwrite,styfile}.c, kana.h,
+	mendex.doc, mendex.1: Add a new option -I to switch
+	internal buffers for key words to EUC-JP or UTF-8.
+	* tests/mendex.test: Add a new test.
+	* tests/uni.{idx,dict,ist} (new): Test input.
+	* tests/uni.ind (new): Expected output.
+	* Makefile.am: Adapted.
+
 2014-05-10  TANAKA Takuji  <KXD02663@nifty.ne.jp>
 
 	* mendex.1: Fix usage.
diff --git a/Build/source/texk/mendexk/Makefile.am b/Build/source/texk/mendexk/Makefile.am
index b9f1dafa102..f27d2fb337c 100644
--- a/Build/source/texk/mendexk/Makefile.am
+++ b/Build/source/texk/mendexk/Makefile.am
@@ -56,6 +56,7 @@ TESTS = tests/mendex.test
 EXTRA_DIST += $(TESTS)
 
 ## tests/mendex.test
-EXTRA_DIST += tests/foo.idx tests/foo.ind
-DISTCLEANFILES = foo.*
+EXTRA_DIST += tests/foo.idx tests/foo.ind \
+	tests/uni.idx tests/uni.dict tests/uni.ist tests/uni.ind
+ (%s)DISTCLEANFILES = foo.* uni.*
 
diff --git a/Build/source/texk/mendexk/convert.c b/Build/source/texk/mendexk/convert.c
index 4d6aff618a2..eae28c970e9 100644
--- a/Build/source/texk/mendexk/convert.c
+++ b/Build/source/texk/mendexk/convert.c
@@ -3,6 +3,8 @@
 #include <kpathsea/tex-file.h>
 #include <kpathsea/variable.h>
 #include <ptexenc/ptexenc.h>
+#include <ptexenc/unicode.h>
+#include <ptexenc/unicode-jp.h>
 
 #include "qsort.h"
 
@@ -21,6 +23,11 @@ static int dlines=0,elines=0;
 
 static int dicvalread(const char *filename, struct dictionary *dicval, int line);
 
+#define is_katakana(a)  ((strncmp((a),KATATOP,2)>=0) && (strncmp((a),KATAEND,2)<=0))
+#define is_hiragana(a)  ((strncmp((a),HIRATOP,2)>=0) && (strncmp((a),HIRAEND,2)<=0))
+#define is_onbiki(a)    (strncmp((a),ONBIKI,2)==0)
+#define is_alpha_numeric(a)  ((strncmp((a),SPACE,2)>=0)&&(strncmp((a),ALPHAEND,2)<=0))
+
 /*   initialize kana table   */
 void initkanatable(void)
 {
@@ -192,6 +199,8 @@ int convert(char *buff1, char *buff2)
 {
 	int i=0,j=0,k,l;
 	char errbuff[4096];
+	int chr,wclen;
+	char buff3[3];
 
 	while(1) {
 		if (buff1[i]=='\0') {
@@ -199,6 +208,28 @@ int convert(char *buff1, char *buff2)
 			break;
 		}
 		else {
+			if ((unsigned char)buff1[i]<0x80)
+				wclen=1;
+			else if (is_internalUPTEX()) {  /* convert a character from UTF8 to EUC */
+				wclen=multibytelen((unsigned char)buff1[i]);
+				if (wclen<0) {
+					verb_printf(efp,"\nWarning: Illegal lead byte 0x%x in UTF-8.", (unsigned char)buff1[i]);
+					i++;
+					continue;
+				}
+				chr = UCS2toJIS(UTF8StoUCS(&buff1[i]));
+				if (chr==0) chr = 0xffff; /* conversion failed */
+				chr |= 0x8080;
+				buff3[0]=BYTE3(chr);
+				buff3[1]=BYTE4(chr);
+				buff3[2]='\0';
+			} else {
+				wclen=2;
+				buff3[0]=buff1[i];
+				buff3[1]=buff1[i+1];
+				buff3[2]='\0';
+			}
+
 			if ((buff1[i]>='a')&&(buff1[i]<='z')) {
 				buff2[j]=buff1[i]-32;
 				i++;
@@ -215,41 +246,41 @@ int convert(char *buff1, char *buff2)
 				j++;
 			}
 
-			else if ((strncmp(&buff1[i],KATATOP,2)>=0)&&(strncmp(&buff1[i],KATAEND,2)<=0)) {
+			else if (is_katakana(buff3)) {
 /*   katakana   */
 				for (k=0;k<strlen(katakana);k+=2) {
-					if (strncmp(&buff1[i],&katakana[k],2)==0) {
+					if (strncmp(buff3,&katakana[k],2)==0) {
 						strncpy(&buff2[j],&kanatable[k],2);
 						goto MATCH1;
 					}
 				}
-				sprintf(errbuff,"\nError: %s is bad katakana ",&buff1[i]);
+				sprintf(errbuff,"\nError: %s is bad katakana ",buff3);
 				fputs(errbuff,efp);
 				if (efp!=stderr) fputs(errbuff,stderr);
 				return -1;
 MATCH1:
-				i+=2;
+				i+=wclen;
 				j+=2;
 			}
 
-			else if ((strncmp(&buff1[i],HIRATOP,2)>=0)&&(strncmp(&buff1[i],HIRAEND,2)<=0)) {
+			else if (is_hiragana(buff3)) {
 /*   hiragana   */
 				for (k=0;k<strlen(hiragana);k+=2) {
-					if (strncmp(&buff1[i],&hiragana[k],2)==0) {
+					if (strncmp(buff3,&hiragana[k],2)==0) {
 						strncpy(&buff2[j],&kanatable[k],2);
 						goto MATCH2;
 					}
 				}
-				sprintf(errbuff,"\nError: %s is bad hiragana ",&buff1[i]);
+				sprintf(errbuff,"\nError: %s is bad hiragana ",buff3);
 				fputs(errbuff,efp);
 				if (efp!=stderr) fputs(errbuff,stderr);
 				return -1;
 MATCH2:
-				i+=2;
+				i+=wclen;
 				j+=2;
 			}
 
-			else if (strncmp(&buff1[i],ONBIKI,2)==0) {
+			else if (is_onbiki(buff3)) {
 /*   onbiki   */
 				if (j>=2) {
 					for (k=0;k<20;k+=2) {
@@ -266,25 +297,26 @@ MATCH2:
 				if (efp!=stderr) fputs(errbuff,stderr);
 				return -1;
 MATCH3:
-				i+=2;
+				i+=wclen;
 				j+=2;
 			}
 
 			else if ((unsigned char)buff1[i]>=0x80) {
-				if ((strncmp(&buff1[i],SPACE,2)>=0)&&(strncmp(&buff1[i],ALPHAEND,2)<=0)) {
+				if (is_alpha_numeric(buff3)) {
 /*   alpha-numeric,symbols   */
 					for (k=0;k<strlen(symboltable);k+=2) {
-						if (strncmp(&buff1[i],&symboltable[k],2)==0) {
+						if (strncmp(buff3,&symboltable[k],2)==0) {
 							buff2[j]=k/2+0x20;
 							if ((buff2[j]>='a')&&(buff2[j]<='z')) buff2[j]-=32;
+							i+=wclen;
 							j++;
-							i+=2;
 							break;
 						}
 					}
 					if (k==strlen(symboltable)) {
-						buff2[j++]=buff1[i++];
-						buff2[j++]=buff1[i++];
+						i+=wclen;
+						buff2[j++]=buff3[0];
+						buff2[j++]=buff3[1];
 					}
 				}
 
@@ -312,8 +344,9 @@ MATCH3:
 					if (((k==dlines)&&(elines==0))||((k==elines)&&(elines!=0))) {
 						if (force==1) {
 /*   forced convert   */
-							buff2[j++]=buff1[i++];
-							buff2[j++]=buff1[i++];
+							i+=wclen;
+							buff2[j++]=buff3[0];
+							buff2[j++]=buff3[1];
 						}
 						else {
 							sprintf(errbuff,"\nError: %s is no entry in dictionary file ",&buff1[i]);
diff --git a/Build/source/texk/mendexk/fread.c b/Build/source/texk/mendexk/fread.c
index e183bca11a4..8ff2eb357f6 100644
--- a/Build/source/texk/mendexk/fread.c
+++ b/Build/source/texk/mendexk/fread.c
@@ -207,9 +207,17 @@ LOOP:
 			wbuff[k]=buff[j];
 			if (buff[j]!=escape) esc=0;
 			if ((unsigned char)buff[j]>=0x80) {
-				wbuff[k+1]=buff[j+1];
-				j++;
-				k++;
+				int len = multibytelen((unsigned char)buff[j]);
+				if (len<0) {
+					verb_printf(efp,"\nWarning: Illegal input of lead byte 0x%x in UTF-8.",(unsigned char)buff[j]);
+					k--;
+					continue;
+				}
+				while(--len) {
+					wbuff[k+1]=buff[j+1];
+					j++;
+					k++;
+				}
 			}
 		}
 		ind[i].words=indent+1;
diff --git a/Build/source/texk/mendexk/fwrite.c b/Build/source/texk/mendexk/fwrite.c
index 77176ec8940..516e772db76 100644
--- a/Build/source/texk/mendexk/fwrite.c
+++ b/Build/source/texk/mendexk/fwrite.c
@@ -4,6 +4,8 @@
 
 #include <kpathsea/tex-file.h>
 #include <ptexenc/ptexenc.h>
+#include <ptexenc/unicode.h>
+#include <ptexenc/unicode-jp.h>
 
 #include "exkana.h"
 #include "exvar.h"
@@ -23,6 +25,25 @@ static void crcheck(char *lbuff, FILE *fp);
 #define BUFFERLEN 4096
 #define TAIL_LEN(x) ((x)+strlen(x)), (BUFFERLEN-strlen(x))
 
+void fprint_euc_char(FILE *fp, const char a, const char b)
+{
+    if (is_internalUPTEX()) {  /* convert a character from EUC to UTF8 */
+	int k = 0;
+	unsigned char str[5];
+	int chr = (unsigned char)a<<8 | (unsigned char)b;
+	chr = (chr==0xffff) ? U_REPLACEMENT_CHARACTER : JIStoUCS2(chr & 0x7f7f);
+	chr = UCStoUTF8(chr);
+	/* if (BYTE1(chr) != 0) str[k++] = BYTE1(chr); */  /* do not happen */
+	if (BYTE2(chr) != 0) str[k++] = BYTE2(chr);
+	if (BYTE3(chr) != 0) str[k++] = BYTE3(chr);
+	                     str[k++] = BYTE4(chr);
+	                     str[k++] = '\0';
+	fprintf(fp,"%s",str);
+    }
+    else
+	fprintf(fp,"%c%c",a,b);
+}
+
 /*   fprintf with convert kanji code   */
 int fprintf2(FILE *fp, const char *format, ...)
 {
@@ -72,6 +93,7 @@ void indwrite(char *filename, struct index *ind, int pagenum)
 	int i,j,hpoint=0;
 	char datama[2048],lbuff[BUFFERLEN];
 	FILE *fp;
+	int conv_euc_to_euc;
 
 	if (filename[0]!='\0' && kpse_out_name_ok(filename)) fp=fopen(filename,"wb");
 	else {
@@ -81,7 +103,10 @@ void indwrite(char *filename, struct index *ind, int pagenum)
 #endif
 	}
 
+	conv_euc_to_euc = is_internalUPTEX() ? 1 : 0;
+	if (conv_euc_to_euc) set_enc_string(NULL, "euc");
 	convert(atama,datama);
+	if (conv_euc_to_euc) set_enc_string(NULL, "uptex");
 	fputs(preamble,fp);
 
 	if (fpage>0) {
@@ -118,13 +143,13 @@ void indwrite(char *filename, struct index *ind, int pagenum)
 					fputs(lethead_prefix,fp);
 					for (j=hpoint;j<(strlen(datama)/2);j++) {
 						if ((unsigned char)ind[i].dic[0][1]<(unsigned char)datama[j*2+1]) {
-							fprintf(fp,"%c%c",atama[(j-1)*2],atama[(j-1)*2+1]);
+							fprint_euc_char(fp,atama[(j-1)*2],atama[(j-1)*2+1]);
 							hpoint=j;
 							break;
 						}
 					}
 					if (j==(strlen(datama)/2)) {
-						fprintf(fp,"%c%c",atama[(j-1)*2],atama[(j-1)*2+1]);
+						fprint_euc_char(fp,atama[(j-1)*2],atama[(j-1)*2+1]);
 					}
 					fputs(lethead_suffix,fp);
 				}
@@ -187,7 +212,7 @@ void indwrite(char *filename, struct index *ind, int pagenum)
 					fputs(group_skip,fp);
 					if (lethead_flag!=0) {
 						fputs(lethead_prefix,fp);
-						fprintf(fp,"%c%c",atama[(j-1)*2],atama[(j-1)*2+1]);
+						fprint_euc_char(fp,atama[(j-1)*2],atama[(j-1)*2+1]);
 						fputs(lethead_suffix,fp);
 					}
 				}
@@ -196,7 +221,7 @@ void indwrite(char *filename, struct index *ind, int pagenum)
 					fputs(group_skip,fp);
 					if (lethead_flag!=0) {
 						fputs(lethead_prefix,fp);
-						fprintf(fp,"%c%c",ind[i].dic[0][0],ind[i].dic[0][1]);
+						fprint_euc_char(fp,ind[i].dic[0][0],ind[i].dic[0][1]);
 						fputs(lethead_suffix,fp);
 					}
 				}
diff --git a/Build/source/texk/mendexk/kana.h b/Build/source/texk/mendexk/kana.h
index 7ad88383b9e..34c17016a1d 100644
--- a/Build/source/texk/mendexk/kana.h
+++ b/Build/source/texk/mendexk/kana.h
@@ -13,6 +13,19 @@ char AKASATANA[]={
 	0xa4,0xef, /* わ */
 	0
 };
+char AKASATANAutf8[]={
+	0xe3,0x81,0x82, /* あ */
+	0xe3,0x81,0x8b, /* か */
+	0xe3,0x81,0x95, /* さ */
+	0xe3,0x81,0x9f, /* た */
+	0xe3,0x81,0xaa, /* な */
+	0xe3,0x81,0xaf, /* は */
+	0xe3,0x81,0xbe, /* ま */
+	0xe3,0x82,0x84, /* や */
+	0xe3,0x82,0x89, /* ら */
+	0xe3,0x82,0x8f, /* わ */
+	0
+};
 char *akasatana;
 
 char AIUEO[]={
diff --git a/Build/source/texk/mendexk/main.c b/Build/source/texk/mendexk/main.c
index 13cda874e7e..52a093e9353 100644
--- a/Build/source/texk/mendexk/main.c
+++ b/Build/source/texk/mendexk/main.c
@@ -20,6 +20,7 @@ char *styfile,*idxfile[256],indfile[256],*dicfile,logfile[256];
 KpathseaSupportInfo kp_ist,kp_dict;
 
 #define VERSION "version 2.6f [14-Aug-2009]"
+#define TL_VERSION "TeX Live 2014"
 
 int main(int argc, char **argv)
 {
@@ -27,7 +28,7 @@ int main(int argc, char **argv)
 	const char *envbuff;
 	char *p;
 
-	enable_UPTEX (false); /* disable */
+	enable_UPTEX (true); /* enable */
 
 #ifdef WIN32
 	_setmaxstdio(2048);
@@ -176,25 +177,39 @@ int main(int argc, char **argv)
 #endif
 
 			case 'U':
-				set_enc_string("UTF8", NULL);
+				set_enc_string("UTF8", "uptex");
+				break;
+
+
+			case 'I':
+				if ((argv[i][2]=='\0')&&(i+1<argc)) {
+					i++;
+					if (strcmp(argv[i],"euc")==0) set_enc_string(NULL, "euc");
+					else if (strcmp(argv[i],"utf8")==0) set_enc_string(NULL, "uptex");
+				}
+				else {
+					if (strcmp(&argv[i][2],"euc")==0) set_enc_string(NULL, "euc");
+					else if (strcmp(&argv[i][2],"utf8")==0) set_enc_string(NULL, "uptex");
+				}
 				break;
 
 			default:
-				fprintf(stderr,"mendex - Japanese index processor, %s (%s).\n",VERSION, get_enc_string());
+				fprintf(stderr,"mendex - Japanese index processor, %s (%s) (%s).\n",VERSION, get_enc_string(), TL_VERSION);
 				fprintf(stderr," Copyright 2009 ASCII MEDIA WORKS.(ptex-staff@ml.asciimw.jp)\n");
 				fprintf(stderr,"usage:\n");
 				fprintf(stderr,"%% mendex [-ilqrcgfEJS"
 #ifdef WIN32
 					       "T"
 #endif
-					       "U] [-s sty] [-d dic] [-o ind] [-t log] [-p no] [idx0 idx1 ...]\n");
+					       "U] [-s sty] [-d dic] [-o ind] [-t log] [-p no] [-I enc] [idx0 idx1 ...]\n");
 				fprintf(stderr,"options:\n");
 				fprintf(stderr,"-i      use stdin as the input file.\n");
 				fprintf(stderr,"-l      use letter ordering.\n");
 				fprintf(stderr,"-q      quiet mode.\n");
 				fprintf(stderr,"-r      disable implicit page formation.\n");
 				fprintf(stderr,"-c      compress blanks. (ignore leading and trailing blanks.)\n");
-				fprintf(stderr,"-g      make Japanese index head <%s>.\n", AKASATANA);
+				fprintf(stderr,"-g      make Japanese index head <%s>.\n",
+							is_internalUPTEX() ? AKASATANAutf8 : AKASATANA);
 				fprintf(stderr,"-f      force to output kanji.\n");
 				fprintf(stderr,"-s sty  take sty as style file.\n");
 				fprintf(stderr,"-d dic  take dic as dictionary file.\n");
@@ -208,6 +223,7 @@ int main(int argc, char **argv)
 				fprintf(stderr,"-T      ShiftJIS terminal.\n");
 #endif
 				fprintf(stderr,"-U      UTF-8 mode.\n");
+				fprintf(stderr,"-I enc  internal encoding for keywords (enc: euc or utf8).\n");
 				fprintf(stderr,"idx...  input files.\n");
 				exit(0);
 				break;
@@ -265,12 +281,12 @@ int main(int argc, char **argv)
 	}
 
 	if (strcmp(argv[0],"makeindex")==0) {
-		verb_printf(efp,"This is Not `MAKEINDEX\', But `MENDEX\' %s (%s).\n",
-			    VERSION, get_enc_string());
+		verb_printf(efp,"This is Not `MAKEINDEX\', But `MENDEX\' %s (%s) (%s).\n",
+			    VERSION, get_enc_string(), TL_VERSION);
 	}
 	else {
-		verb_printf(efp,"This is mendex %s (%s).\n",
-			    VERSION, get_enc_string());
+		verb_printf(efp,"This is mendex %s (%s) (%s).\n",
+			    VERSION, get_enc_string(), TL_VERSION);
 	}
 
 /*   init kanatable   */
diff --git a/Build/source/texk/mendexk/mendex.1 b/Build/source/texk/mendexk/mendex.1
index 48360700e90..c19f1cc53b4 100644
--- a/Build/source/texk/mendexk/mendex.1
+++ b/Build/source/texk/mendexk/mendex.1
@@ -4,7 +4,7 @@
 .SH 名称
 mendex \- 索引整形ツール
 .SH 使用法
-\fBmendex\fR [-ilqrcgEJS] [\fB-s\fI sty\fR] [\fB-d\fI dic\fR] [\fB-o\fI ind\fR] [\fB-d\fI dic\fR] [\fB-t\fI log\fR] [\fB-p\fI no\fR] [\fI idx0 idx1 idx2 ...\fR]
+\fBmendex\fR [-ilqrcgEJS] [\fB-s\fI sty\fR] [\fB-d\fI dic\fR] [\fB-o\fI ind\fR] [\fB-d\fI dic\fR] [\fB-t\fI log\fR] [\fB-p\fI no\fR] [\fB-I\fI enc\fR] [\fI idx0 idx1 idx2 ...\fR]
 .SH 解説
 .PP
 \fImendex\fR は文書の索引を作成するツールです。 LaTeX により抽出された索引リストファイル(\fI.idx\fR)を並べ替え、実際の索引のソースファイルの形に整形します。 \fImakeindex\fR と互換性があり、さらに「読み」の扱いの手間を減らすように特殊化されています。
@@ -62,6 +62,9 @@ mendex \- 索引整形ツール
 .TP 10
 \fB-U\fR
 漢字コードをUTF-8に指定します。入力ファイル、出力ファイルともUTF-8として扱います。
+.TP 10
+\fB-I\fI enc\fR
+内部バッファのコードを\fIenc\fR に指定します。\fIenc\fR には \fIeuc\fR(EUC-JP) または \fIutf8\fR(UTF-8)が指定可能です。
 
 
 .SH スタイルファイル
diff --git a/Build/source/texk/mendexk/mendex.doc b/Build/source/texk/mendexk/mendex.doc
index 1a3505df4f8..adae776fef8 100644
--- a/Build/source/texk/mendexk/mendex.doc
+++ b/Build/source/texk/mendexk/mendex.doc
@@ -4,7 +4,7 @@ mendex - 索引整形ツール
 
 使用法
 
-    mendex [-ilqrcgEJS] [-s sty] [-d dic] [-o ind] [-d dic] [-t log] [-p no] [idx0 idx1 idx2 ...]
+    mendex [-ilqrcgEJS] [-s sty] [-d dic] [-o ind] [-d dic] [-t log] [-p no] [-I enc] [idx0 idx1 idx2 ...]
 
 
 解説
@@ -66,6 +66,8 @@ mendex - 索引整形ツール
    -S      漢字コードをShiftJISに指定します。
            入力ファイル、出力ファイルともShiftJISとして扱います。
 
+   -I enc  内部バッファのコードをencに指定します。enc には euc (EUC-JP) または utf8 (UTF-8)が指定可能です。
+
 
 スタイルファイル
 
diff --git a/Build/source/texk/mendexk/styfile.c b/Build/source/texk/mendexk/styfile.c
index 9179a34904c..04189664099 100644
--- a/Build/source/texk/mendexk/styfile.c
+++ b/Build/source/texk/mendexk/styfile.c
@@ -165,8 +165,16 @@ static void convline(char *buff1, int start, char *buff2)
 		}
 		else if (cc==1) {
 			if ((unsigned char)buff1[i]>0x80) {
-				 buff2[j++]=buff1[i++];
-				 buff2[j]=buff1[i];
+				int len = multibytelen((unsigned char)buff1[i]);
+				if (len<0) {
+					verb_printf(efp,"\nWarning: Illegal input of lead byte 0x%x in UTF-8.", (unsigned char)buff1[i]);
+					continue;
+				}
+				while(len--) {
+					buff2[j++]=buff1[i++];
+				}
+				i--;
+				continue;
 			}
 			else if (buff1[i]=='\\') {
 				i++;
diff --git a/Build/source/texk/mendexk/tests/mendex.test b/Build/source/texk/mendexk/tests/mendex.test
index 6afaa17d92d..53988d1eab5 100755
--- a/Build/source/texk/mendexk/tests/mendex.test
+++ b/Build/source/texk/mendexk/tests/mendex.test
@@ -14,3 +14,12 @@ rm -f foo.*
 cat $srcdir/tests/foo.idx | ./mendex >foo.ind2 2>foo.ilg2 \
 	&& diff $srcdir/tests/foo.ind foo.ind2 || exit 1
 
+./mendex -U -d $srcdir/tests/uni.dict -s $srcdir/tests/uni.ist \
+	$srcdir/tests/uni.idx -o uni.ind1 -t uni.ilg1 \
+	&& diff $srcdir/tests/uni.ind uni.ind1 || exit 1
+
+cat $srcdir/tests/uni.idx | \
+	./mendex -U -d $srcdir/tests/uni.dict -s $srcdir/tests/uni.ist \
+	>uni.ind2 2>uni.ilg2 \
+	&& diff $srcdir/tests/uni.ind uni.ind2 || exit 1
+
diff --git a/Build/source/texk/mendexk/tests/uni.dict b/Build/source/texk/mendexk/tests/uni.dict
new file mode 100644
index 00000000000..36343dac17d
--- /dev/null
+++ b/Build/source/texk/mendexk/tests/uni.dict
@@ -0,0 +1,3 @@
+% -*- coding: utf-8 -*-
+�������  �����ｃ��
+羶����  ������������
diff --git a/Build/source/texk/mendexk/tests/uni.idx b/Build/source/texk/mendexk/tests/uni.idx
new file mode 100644
index 00000000000..52a0b8ecea9
--- /dev/null
+++ b/Build/source/texk/mendexk/tests/uni.idx
@@ -0,0 +1,8 @@
+\indexentry{������������������@罍�薹�紊�}{1}
+\indexentry{��������帥�����@���荀�綣�}{1}
+\indexentry{������ゆ�����}{1}
+\indexentry{Mendex}{1}
+\indexentry{Mendex!�����ゃ�潟�鴻����若��}{1}
+\indexentry{Mendex!�����ゃ�����������@---���篏帥�����}{1}
+\indexentry{Mendex!�����ゃ�����������@---���篏帥�����!�����������������吾��@---��ラ��膩�}{1}
+\indexentry{Mendex!�����ゃ�����������@---���篏帥�����!��������������吾��@---綽����膩�}{1}
diff --git a/Build/source/texk/mendexk/tests/uni.ind b/Build/source/texk/mendexk/tests/uni.ind
new file mode 100644
index 00000000000..aa142686417
--- /dev/null
+++ b/Build/source/texk/mendexk/tests/uni.ind
@@ -0,0 +1,25 @@
+\begin{theindex}
+
+\centerline{\bfseries ��������� M ���������}\par\nobreak
+  \item Mendex\leaders\hbox{$\cdot$}\hfill 1
+    \subitem �����ゃ�潟�鴻����若��\leaders\hbox{$\cdot$}\hfill 1
+    \subitem ---���篏帥�����\leaders\hbox{$\cdot$}\hfill 1
+      \subsubitem ---綽����膩�\leaders\hbox{$\cdot$}\hfill 1
+      \subsubitem ---��ラ��膩�\leaders\hbox{$\cdot$}\hfill 1
+
+  \indexspace
+
+\centerline{\bfseries ��������� ��� ���������}\par\nobreak
+  \item ���荀�綣�\leaders\hbox{$\cdot$}\hfill 1
+
+  \indexspace
+
+\centerline{\bfseries ��������� ��� ���������}\par\nobreak
+  \item ������ゆ�����\leaders\hbox{$\cdot$}\hfill 1
+
+  \indexspace
+
+\centerline{\bfseries ��������� ��� ���������}\par\nobreak
+  \item 罍�薹�紊�\leaders\hbox{$\cdot$}\hfill 1
+
+\end{theindex}
diff --git a/Build/source/texk/mendexk/tests/uni.ist b/Build/source/texk/mendexk/tests/uni.ist
new file mode 100644
index 00000000000..095533d1fb3
--- /dev/null
+++ b/Build/source/texk/mendexk/tests/uni.ist
@@ -0,0 +1,10 @@
+% -*- coding: utf-8 -*-
+
+delim_0 "\\leaders\\hbox{$\\cdot$}\\hfill "
+delim_1 "\\leaders\\hbox{$\\cdot$}\\hfill "
+delim_2 "\\leaders\\hbox{$\\cdot$}\\hfill "
+
+lethead_flag 1
+lethead_prefix "\n\\centerline{\\bfseries ��������� "
+lethead_suffix  " ���������}\\par\\nobreak"
+