From 8ab85c695504819695b18486b80383412d4e9177 Mon Sep 17 00:00:00 2001 From: Takuji Tanaka Date: Sat, 22 Feb 2020 06:05:06 +0000 Subject: upmendex v0.54 git-svn-id: svn://tug.org/texlive/trunk@53866 c570f23f-e606-0410-a88d-b1316a301751 --- Build/source/texk/upmendex/ChangeLog | 10 ++++++++++ Build/source/texk/upmendex/configure | 22 +++++++++++----------- Build/source/texk/upmendex/configure.ac | 2 +- Build/source/texk/upmendex/convert.c | 3 ++- Build/source/texk/upmendex/fwrite.c | 19 +++++++++++++++++-- Build/source/texk/upmendex/main.c | 4 ++-- Build/source/texk/upmendex/sort.c | 14 ++++++++++++-- 7 files changed, 55 insertions(+), 19 deletions(-) (limited to 'Build/source/texk/upmendex') diff --git a/Build/source/texk/upmendex/ChangeLog b/Build/source/texk/upmendex/ChangeLog index af3ba3f8fb6..6c0e2aaef79 100644 --- a/Build/source/texk/upmendex/ChangeLog +++ b/Build/source/texk/upmendex/ChangeLog @@ -1,3 +1,13 @@ +2020-02-22 TANAKA Takuji + + * version 0.54 Beta version. + * configure.ac, main.c, COPYRIGHT: Bump version + and copyright year. + * {convert,fwrite,sort}.c: + Support Small Kana Extension in Unicode 12.0. + Add CJK Unified Ideographs Extension G. + Update samples. + 2020-01-16 Karl Berry * configure.ac, diff --git a/Build/source/texk/upmendex/configure b/Build/source/texk/upmendex/configure index aa97139d312..55a19d16496 100755 --- a/Build/source/texk/upmendex/configure +++ b/Build/source/texk/upmendex/configure @@ -1,6 +1,6 @@ #! /bin/sh # Guess values for system-dependent variables and create Makefiles. -# Generated by GNU Autoconf 2.69 for upmendex (TeX Live) 0.53. +# Generated by GNU Autoconf 2.69 for upmendex (TeX Live) 0.54. # # # Copyright (C) 1992-1996, 1998-2012 Free Software Foundation, Inc. @@ -587,8 +587,8 @@ MAKEFLAGS= # Identity of this package. PACKAGE_NAME='upmendex (TeX Live)' PACKAGE_TARNAME='upmendex--tex-live-' -PACKAGE_VERSION='0.53' -PACKAGE_STRING='upmendex (TeX Live) 0.53' +PACKAGE_VERSION='0.54' +PACKAGE_STRING='upmendex (TeX Live) 0.54' PACKAGE_BUGREPORT='' PACKAGE_URL='' @@ -1345,7 +1345,7 @@ if test "$ac_init_help" = "long"; then # Omit some internal or obsolete options to make the list less imposing. # This message is too long to be a string in the A/UX 3.1 sh. cat <<_ACEOF -\`configure' configures upmendex (TeX Live) 0.53 to adapt to many kinds of systems. +\`configure' configures upmendex (TeX Live) 0.54 to adapt to many kinds of systems. Usage: $0 [OPTION]... [VAR=VALUE]... @@ -1416,7 +1416,7 @@ fi if test -n "$ac_init_help"; then case $ac_init_help in - short | recursive ) echo "Configuration of upmendex (TeX Live) 0.53:";; + short | recursive ) echo "Configuration of upmendex (TeX Live) 0.54:";; esac cat <<\_ACEOF @@ -1541,7 +1541,7 @@ fi test -n "$ac_init_help" && exit $ac_status if $ac_init_version; then cat <<\_ACEOF -upmendex (TeX Live) configure 0.53 +upmendex (TeX Live) configure 0.54 generated by GNU Autoconf 2.69 Copyright (C) 2012 Free Software Foundation, Inc. @@ -2474,7 +2474,7 @@ cat >config.log <<_ACEOF This file contains any messages produced by compilers while running configure, to aid debugging if configure makes a mistake. -It was created by upmendex (TeX Live) $as_me 0.53, which was +It was created by upmendex (TeX Live) $as_me 0.54, which was generated by GNU Autoconf 2.69. Invocation command line was $ $0 $@ @@ -8673,7 +8673,7 @@ fi # Define the identity of the package. PACKAGE='upmendex--tex-live-' - VERSION='0.53' + VERSION='0.54' # Some tools Automake needs. @@ -18202,7 +18202,7 @@ Usage: $0 [OPTIONS] Report bugs to ." lt_cl_version="\ -upmendex (TeX Live) config.lt 0.53 +upmendex (TeX Live) config.lt 0.54 configured by $0, generated by GNU Autoconf 2.69. Copyright (C) 2011 Free Software Foundation, Inc. @@ -20342,7 +20342,7 @@ cat >>$CONFIG_STATUS <<\_ACEOF || ac_write_fail=1 # report actual input values of CONFIG_FILES etc. instead of their # values after options handling. ac_log=" -This file was extended by upmendex (TeX Live) $as_me 0.53, which was +This file was extended by upmendex (TeX Live) $as_me 0.54, which was generated by GNU Autoconf 2.69. Invocation command line was CONFIG_FILES = $CONFIG_FILES @@ -20408,7 +20408,7 @@ _ACEOF cat >>$CONFIG_STATUS <<_ACEOF || ac_write_fail=1 ac_cs_config="`$as_echo "$ac_configure_args" | sed 's/^ //; s/[\\""\`\$]/\\\\&/g'`" ac_cs_version="\\ -upmendex (TeX Live) config.status 0.53 +upmendex (TeX Live) config.status 0.54 configured by $0, generated by GNU Autoconf 2.69, with options \\"\$ac_cs_config\\" diff --git a/Build/source/texk/upmendex/configure.ac b/Build/source/texk/upmendex/configure.ac index f4596fb0ca5..c95d31520be 100644 --- a/Build/source/texk/upmendex/configure.ac +++ b/Build/source/texk/upmendex/configure.ac @@ -8,7 +8,7 @@ dnl This file is free software; the copyright holder dnl gives unlimited permission to copy and/or distribute it, dnl with or without modifications, as long as this notice is preserved. dnl -AC_INIT([upmendex (TeX Live)], [0.53]) +AC_INIT([upmendex (TeX Live)], [0.54]) AC_PREREQ([2.63]) AC_CONFIG_SRCDIR([main.c]) AC_CONFIG_AUX_DIR([../../build-aux]) diff --git a/Build/source/texk/upmendex/convert.c b/Build/source/texk/upmendex/convert.c index 72e7cb5b783..d02b05bf4d4 100644 --- a/Build/source/texk/upmendex/convert.c +++ b/Build/source/texk/upmendex/convert.c @@ -217,7 +217,8 @@ int convert(UChar *buff1, UChar *buff2) ||is_jpn_kana(buff3)||is_kor_hngl(buff3)||is_zhuyin(buff3) ||is_comb_diacritical_mark(buff3)) { buff2[j]=buff3[0]; - i++; + if (wclen==2) buff2[j+1]=buff3[1]; + i+=wclen; j+=wclen; } diff --git a/Build/source/texk/upmendex/fwrite.c b/Build/source/texk/upmendex/fwrite.c index 6736f56e5a2..380e1cbf37a 100644 --- a/Build/source/texk/upmendex/fwrite.c +++ b/Build/source/texk/upmendex/fwrite.c @@ -649,6 +649,7 @@ static void index_normalize(UChar *istr, UChar *ini, int *chset) { int k, len, hi, lo, mi; UChar ch,src[2],dest[8],strX[4],strY[4],strZ[4]; + UChar32 c32; UErrorCode perr; UCollationResult order; static int hanzi_mode=0; @@ -668,8 +669,22 @@ static void index_normalize(UChar *istr, UChar *ini, int *chset) ini[0]=extkanatable[ch-EXKANATOP]; return; } - else if (ch==0x309F) { ini[0]=0x30E8; return; } /* HIRAGANA YORI */ - else if (ch==0x30FF) { ini[0]=0x30B3; return; } /* KATAKANA KOTO */ + else if (ch==0x309F) { ini[0]=0x30E8; return; } /* HIRAGANA YORI -> ヨ */ + else if (ch==0x30FF) { ini[0]=0x30B3; return; } /* KATAKANA KOTO -> コ */ + else if (is_jpn_kana(istr)==2) { + c32=U16_GET_SUPPLEMENTARY(istr[0],istr[1]); + switch (c32) { + case 0x1B150: case 0x1B164: + ini[0]=0x30F0; break; /* ヰ */ + case 0x1B151: case 0x1B165: + ini[0]=0x30F1; break; /* ヱ */ + case 0x1B152: case 0x1B166: + ini[0]=0x30F2; break; /* ヲ */ + case 0x1B167: default: + ini[0]=0x30F3; break; /* ン */ + } + return; + } else if (is_kor_hngl(&ch)) { if ((ch>=0xAC00)&&(ch<=0xD7AF)) { /* Hangul Syllables */ ch=(ch-0xAC00)/(21*28)+CHOSEONG_KIYEOK; /* convert to Hangul Jamo, Initial consonants */ diff --git a/Build/source/texk/upmendex/main.c b/Build/source/texk/upmendex/main.c index 3d7f0f59459..c7d8106b537 100644 --- a/Build/source/texk/upmendex/main.c +++ b/Build/source/texk/upmendex/main.c @@ -19,7 +19,7 @@ char *styfile,*idxfile[256],*indfile,*dicfile,*logfile; #endif KpathseaSupportInfo kp_ist,kp_dict; -#define VERSION "version 0.53" +#define VERSION "version 0.54" int main(int argc, char **argv) { @@ -156,7 +156,7 @@ int main(int argc, char **argv) u_getVersion(icuVersion); u_versionToString(icuVersion, icu_version); fprintf(stderr,"upmendex - index processor, %s (%s).\n",VERSION, TL_VERSION); - fprintf(stderr," Copyright 2009 ASCII MEDIA WORKS, 2015-2019 TANAKA Takuji\n"); + fprintf(stderr," Copyright 2009 ASCII MEDIA WORKS, 2015-2020 TANAKA Takuji\n"); fprintf(stderr," using ICU version %s\n",icu_version); fprintf(stderr,"usage:\n"); fprintf(stderr,"%% upmendex [-ilqrcgf] [-s sty] [-d dic] [-o ind] [-t log] [-p no] [--] [idx0 idx1 ...]\n"); diff --git a/Build/source/texk/upmendex/sort.c b/Build/source/texk/upmendex/sort.c index 5d11a0727c0..15272dc8594 100644 --- a/Build/source/texk/upmendex/sort.c +++ b/Build/source/texk/upmendex/sort.c @@ -366,9 +366,18 @@ int is_numeric(UChar *c) int is_jpn_kana(UChar *c) { + UChar32 c32; + if ((*c>=0x3040)&&(*c<=0x30FF)) return 1; /* Hiragana, Katakana */ else if ((*c>=0x31F0)&&(*c<=0x31FF)) return 1; /* Katakana Phonetic Extensions */ - else return 0; + + if (is_surrogate_pair(c)) { + c32=U16_GET_SUPPLEMENTARY(*c,*(c+1)); + if ((c32>=0x1B130) && (c32<=0x1B16F)) return 2; /* Small Kana Extensions */ + } + return 0; + /* ICU 65 does not seem to support + "Kana Supplement" and "Kana Extended-A" yet. (2020/02/16) */ } int is_kor_hngl(UChar *c) @@ -399,7 +408,8 @@ int is_hanzi(UChar *c) if (is_surrogate_pair(c)) { c32=U16_GET_SUPPLEMENTARY(*c,*(c+1)); if ((c32>=0x20000) && /* CJK Unified Ideographs Extension B,C,D,E,F */ - (c32<=0x2FA1F)) return 2; /* CJK Compatibility Ideographs Supplement */ + /* CJK Compatibility Ideographs Supplement */ + (c32<=0x3134F)) return 2; /* CJK Unified Ideographs Extension G */ } return 0; } -- cgit v1.2.3