system/unac: Updated for version 1.8.0_9.

Signed-off-by: B. Watson <yalhcru@gmail.com> Signed-off-by: Willy Sudiarto Raharjo <willysr@slackbuilds.org>
2021-08-20 14:59:31 -04:00 · 2021-08-20 14:59:31 -04:00 · 78298c73d1
parent 85ef26ae28
commit 78298c73d1
6 changed files with 53 additions and 79 deletions
--- a/system/unac/README
+++ b/system/unac/README
@ -1,5 +1,7 @@
-unac is a C library and command that removes accents from a string
-It provides a command line interface that removes accents from a
-input flow or a string given in argument (unaccent command).
-In the library function and the command, the charset of the input
-is specified as an argument.
+unac (remove accents from text)
+
+unac is a C library (libunac) and command (unaccent) that removes
+accents from a string of text; it replaces accented characters in the
+given character set with their unaccented (ASCII) equivalents.
+
+See also: perl/perl-text-unaccent, for a Perl module based on this.
--- a/system/unac/oe_ligature.diff
+++ b/system/unac/oe_ligature.diff
@ -1,11 +1,16 @@
 diff -Naur unac-1.8.0.orig/unaccent-local-map.perl unac-1.8.0.orig.patched/unaccent-local-map.perl
 --- unac-1.8.0.orig/unaccent-local-map.perl	2003-10-13 06:24:25.000000000 -0400
-+++ unac-1.8.0.orig.patched/unaccent-local-map.perl	2015-10-26 17:01:49.000000000 -0400
-@@ -23,5 +23,7 @@
+++ unac-1.8.0.orig.patched/unaccent-local-map.perl	2021-08-20 14:13:53.224474419 -0400
+@@ -23,5 +23,12 @@
 %::localmap = (
     "LATIN SMALL LETTER AE" => [ "LATIN SMALL LETTER A", "LATIN SMALL LETTER E" ],
     "LATIN CAPITAL LETTER AE" => [ "LATIN CAPITAL LETTER A", "LATIN CAPITAL LETTER E" ],
 +    "LATIN SMALL LIGATURE OE" => [ "LATIN SMALL LETTER O", "LATIN SMALL LETTER E" ],
 +    "LATIN CAPITAL LIGATURE OE" => [ "LATIN CAPITAL LETTER O", "LATIN CAPITAL LETTER E" ],
+    "LATIN SMALL LETTER SHARP S" => [ "LATIN SMALL LETTER S", "LATIN SMALL LETTER S" ],
+    "LATIN SMALL LETTER HV" => [ "LATIN SMALL LETTER H", "LATIN SMALL LETTER V" ],
 		 );
 
+# 20210820 bkw: Unicode standard 4.0.0 doesn't define these:
+#    "LATIN CAPITAL LETTER HV" => [ "LATIN CAPITAL LETTER H", "LATIN CAPITAL LETTER V" ],
+#    "LATIN CAPITAL LETTER SHARP S" => [ "LATIN CAPITAL LETTER S", "LATIN CAPITAL LETTER S" ],
--- a/system/unac/size_t.diff
+++ b/system/unac/size_t.diff
@ -1,45 +0,0 @@
-diff -r fd7eb0d70970 unac-1.8.0/unac.c
--- a/unac-1.8.0/unac.c	Fri Sep 17 10:35:14 2010 +0200
-+++ b/unac-1.8.0/unac.c	Fri Sep 17 10:36:07 2010 +0200
-@@ -13873,9 +13873,9 @@
-     *out_lengthp = 0;
-   } else {
-     char* utf16 = 0;
-    int utf16_length = 0;
-+    size_t utf16_length = 0;
-     char* utf16_unaccented = 0;
-    int utf16_unaccented_length = 0;
-+    size_t utf16_unaccented_length = 0;
-   
-     if(convert(charset, utf16be(), in, in_length, &utf16, &utf16_length) < 0) {
-       return -1;
-diff -r fd7eb0d70970 unac-1.8.0/unaccent.c
--- a/unac-1.8.0/unaccent.c	Fri Sep 17 10:35:14 2010 +0200
-+++ b/unac-1.8.0/unaccent.c	Fri Sep 17 10:40:34 2010 +0200
-@@ -90,7 +90,7 @@
-     const char* charset = argv[optind++];
- 
-     char* unaccented = 0;
-    int unaccented_length = 0;
-+    size_t unaccented_length = 0;
- 
-     if(optind >= argc) {
- #define BUFFER_SIZE 10240
-@@ -101,7 +101,7 @@
- 	  perror("");
- 	  exit(1);
- 	}
-	printf("%.*s", unaccented_length, unaccented);
-+	printf("%.*s", (int)unaccented_length, unaccented);
-       }
-     } else {
-       const char* string = argv[optind++];
-@@ -114,7 +114,7 @@
- 
-       if(debug_level > UNAC_DEBUG_NONE)
- 	fprintf(stderr, "unaccented version is ");
-      printf("%.*s\n", unaccented_length, unaccented);
-+      printf("%.*s\n", (int)unaccented_length, unaccented);
- 
-       if(optind < argc) {
- 	const char* expected = argv[optind++];
--- a/system/unac/slack-desc
+++ b/system/unac/slack-desc
@ -8,11 +8,11 @@
    |-----handy-ruler------------------------------------------------------|
 unac: unac (remove accents from text)
 unac:
-unac: unac is a C library and command that removes accents from a string
-unac: It provides a command line interface that removes accents from a
-unac: input flow or a string given in argument (unaccent command).
-unac: In the library function and the command, the charset of the input
-unac: is specified as an argument.
+unac: unac is a C library (libunac) and command (unaccent) that removes
+unac: accents from a string of text; it replaces accented characters in the
+unac: given character set with their unaccented (ASCII) equivalents.
+unac:
+unac:
 unac:
 unac:
 unac:
--- a/system/unac/unac.SlackBuild
+++ b/system/unac/unac.SlackBuild
@ -6,10 +6,19 @@

 # Licensed under the WTFPL. See http://www.wtfpl.net/txt/copying/ for details.

+# 20210819 bkw:
+# - updated to match debian's patchlevel 9.
+# - add a few more ligatures.
+
+# Note to self: don't try to update this for Unicode standard 13.0.0.
+# It'll compile and run, but the code assumes the codepoints will fit
+# in an unsigned short (16 bits), which is no longer true in the new
+# standard.
+
 cd $(dirname $0) ; CWD=$(pwd)

 PRGNAM=unac
-VERSION=${VERSION:-1.8.0}
+VERSION=${VERSION:-1.8.0_9}
 BUILD=${BUILD:-1}
 TAG=${TAG:-_SBo}
 PKGTYPE=${PKGTYPE:-tgz}
@ -22,9 +31,6 @@ if [ -z "$ARCH" ]; then
  esac
 fi

-# If the variable PRINT_PACKAGE_NAME is set, then this script will report what
-# the name of the created package would be, and then exit. This information
-# could be useful to other scripts.
 if [ ! -z "${PRINT_PACKAGE_NAME}" ]; then
  echo "$PRGNAM-$VERSION-$ARCH-$BUILD$TAG.$PKGTYPE"
  exit 0
@ -50,8 +56,13 @@ fi

 set -e

-TARBALL=${PRGNAM}_${VERSION}.orig.tar.gz
-DIRNAME=${PRGNAM}-${VERSION}.orig
+# The original upstream hasn't maintained this since 2004, but
+# the Debian maintainers have been keeping it up to date.
+MAINVER="${VERSION//_*}"  # e.g. 1.8.0_9 => 1.8.0
+DEBVER="${VERSION//*_}"   # e.g. 1.8.0_9 => 9
+TARBALL=${PRGNAM}_${MAINVER}.orig.tar.gz
+DEBTAR=${PRGNAM}_${MAINVER}-${DEBVER}.debian.tar.xz
+DIRNAME=${PRGNAM}-${MAINVER}.orig

 rm -rf $PKG
 mkdir -p $TMP $PKG $OUTPUT
@ -59,21 +70,19 @@ cd $TMP
 rm -rf $DIRNAME
 tar xvf $CWD/$TARBALL
 cd $DIRNAME
+tar xvf $CWD/$DEBTAR
 chown -R root:root .
-find -L . \
- \( -perm 777 -o -perm 775 -o -perm 750 -o -perm 711 -o -perm 555 \
-  -o -perm 511 \) -exec chmod 755 {} \; -o \
- \( -perm 666 -o -perm 664 -o -perm 640 -o -perm 600 -o -perm 444 \
-  -o -perm 440 -o -perm 400 \) -exec chmod 644 {} \;
+find -L .  -perm /111 -a \! -perm 755 -a -exec chmod 755 {} \+ -o \
+        \! -perm /111 -a \! -perm 644 -a -exec chmod 644 {} \+

-# patch taken from http://http.debian.net/debian/pool/main/u/unac/unac_1.8.0-6.diff.gz
-# (yes, it needs -p2)
-patch -p2 < $CWD/size_t.diff
+for i in $( cat debian/patches/series ); do
+  patch -p1 < debian/patches/$i
+done

-# patch by slackbuild author, add support for OE ligature
-patch -p1 < $CWD/oe_ligature.diff
+# patch by slackbuild author, add support for OE ligature, German sharp S,
+# and Hwair.
+patch -p1 < $CWD/ligatures.diff

-touch config.rpath # shut up, autocrap
 autoreconf -if

 CFLAGS="$SLKCFLAGS" \
@ -89,11 +98,12 @@ CXXFLAGS="$SLKCFLAGS" \
  --enable-shared \
  --build=$ARCH-slackware-linux

-# this is needed for the OE ligature also
+# this is needed for the ligatures also
 make -B $PRGNAM.c

 make
 make install-strip DESTDIR=$PKG
+rm -f $PKG/usr/lib$LIBDIRSUFFIX/*.la
 gzip $PKG/usr/man/man?/*.?

 mkdir -p $PKG/usr/doc/$PRGNAM-$VERSION
--- a/system/unac/unac.info
+++ b/system/unac/unac.info
@ -1,8 +1,10 @@
 PRGNAM="unac"
-VERSION="1.8.0"
-HOMEPAGE="http://www.nongnu.org/unac/"
-DOWNLOAD="http://http.debian.net/debian/pool/main/u/unac/unac_1.8.0.orig.tar.gz"
-MD5SUM="8670c920969a199dec37f15c197602bf"
+VERSION="1.8.0_9"
+HOMEPAGE="https://www.nongnu.org/unac/"
+DOWNLOAD="https://deb.debian.org/debian/pool/main/u/unac/unac_1.8.0.orig.tar.gz \
+          https://deb.debian.org/debian/pool/main/u/unac/unac_1.8.0-9.debian.tar.xz"
+MD5SUM="8670c920969a199dec37f15c197602bf \
+        5dcc4e384b015aebb7dd38569c22b6d0"
 DOWNLOAD_x86_64=""
 MD5SUM_x86_64=""
 REQUIRES=""