Collection.
The Unicode::Map8 class implement efficient mapping tables between
8-bit character sets and 16 bit character sets like Unicode. About
170 different mapping tables between various known character sets and
Unicode is distributed with this package. The source of these tables
is the vendor mapping tables provided by Unicode, Inc. and the code
tables in RFC 1345. New maps can easily be installed.
Changes since last packaged version (2.06):
2003-03-10 Gisle Aas <gisle@ActiveState.com>
Release 2.07
Don't let the $^W test get confused by lexical warnings.
0.9.6 20040430 Compilation fixes for Solaris
Compilation fixes for ImageMagick unavailability
0.9.5 20040406 Bug fixes
Compilation with gcc 2.95.x should work again
0.9.4 20040316 Mostly bug fixes
Minimum version required for AbiWord 2.1.1 and later
0.9.3 20040226 Minimum version required for new AbiWord 2.1 plugin
Fixed a few minor bugs
Added psiconv_unicode_from_list and psiconv_unicode_strstr
Much better error and progress reporting in generators
0.9.2.1 20040223 Quick fix for the library number mistake in 0.9.2
0.9.2 20040223 Many bugfixes, among them some critical ones:
* Package now builds without psiconv.conf files installed
UTF8 Psion support might now actually work
-c argument introduced for psiconv program
0.9.1 20040209 Added the psiconv-config program.
Fixed the installation of @includedir@/psiconv/general.h
Added the forgotted psiconv_config_free function.
Removed some small memory leaks.
0.9.0 20040204 New major library version number breaks source compatibility:
* All strings and characters are now stored in UCS-2
* New config structure.
- Call psiconv_config_default and psiconv_config_read at the
start of your program
- Almost all functions have a new psiconv_config parameter
- All former global variables are now in this structure
* Several minor changes in the functions defined in
parse_routines.h and generate_routines.h
* All generate_* routines have now a lev parameter except
generate_file.
* psiconv_sketch_section data representation changes
* New verbosity level PSICONV_VERB_ERROR
Use config files /etc/psiconv.conf and ~/.psiconv.conf
Support automake 1.6, 1.7 and 1.8
Support autoconf-2.50 and up
Support ImageMagick 5.4.x and 5.5.x (API changed once again...)
Support Unicode output in psiconv program (Keita Kawabe)
Don't use features not present in plain sh (Keita Kawabe)
Clean up some automake-related stuff.
Format documentation is now installed too.
Debian build support
Embedded objects in Word are now properly parsed and generated
Sketch file generation is added (stand-alone and as object)
MBM and Clipart file generation is added
Image files other than 2-bits greyscale are parsed and
generated (experimental)
Errors and warnings sanitized
Rewrite of the psiconv program:
* Outputs in UTF8, UCS2, ASCII or Psion encodings
* XHTML target generates strict XHTML using CSS
* HTML4 target generates traditional HTML without CSS
* IMAGE target remains the same
* LATEX and RTF targets are not supported anymore
in the process. (More information on tech-pkg.)
Bump PKGREVISION and BUILDLINK_DEPENDS of all packages using libtool and
installing .la files.
Bump PKGREVISION (only) of all packages depending directly on the above
via a buildlink3 include.
All library names listed by *.la files no longer need to be listed
in the PLIST, e.g., instead of:
lib/libfoo.a
lib/libfoo.la
lib/libfoo.so
lib/libfoo.so.0
lib/libfoo.so.0.1
one simply needs:
lib/libfoo.la
and bsd.pkg.mk will automatically ensure that the additional library
names are listed in the installed package +CONTENTS file.
Also make LIBTOOLIZE_PLIST default to "yes".
BinHex is a format used by Macintosh for transporting Mac files safely
through electronic mail, as short-lined, 7-bit, semi-compressed data
streams. Ths module provides a means of converting those data streams
back into into binary data.
to after GNU_CONFIGURE and USE_LIBTOOL being set.
This fixes the problem where the lt_cv_sys_max_cmd_len was not
added to the CONFIGURE_ENV.
Now that I think about it, I caused this problem when I added that
Linux check...
intended transformation: use "rm" to remove an option, "rmdir" to remove
all options containing a path starting with a given directory name, and
"rename" to rename options to something else.
$Revision: 0.87 $ $Date: 2004/07/24 02:58:27 $
! Jcode.pm
* Document Fix by shigeno@iij-mc.co.jp
Message-Id: <20040708.175523.131609260.shigeno@iij-mc.co.jp>
* getcode() reverted to pre-0.84 because the new algorithm is prone
to false positive cases, reported by inu@kc5.so-net.ne.jp.
Message-Id: <20040723.233005.78712771.inu@mail.design.co.jp>
include:
- renamed crc32 to uulib_crc32, to work around yet another shortcoming
in this oh-so-outdated macosx.
- very short files might not be detected because the header parsing
code might skip them.
- change bracket policy to give priority to (x/y) over (x), and use
the last bracket found otherwise.
- part numbers at the end of the subject are now recognized.
more recent PHP anymore, so use USE_BUILDIN.iconv test to select
the appropriate iconv parameter value
fixes compilation on NetBSD 1.6.x and other systems without builtin
iconv, and thus PR pkg/26212 and PR pkg/26213
package behaviour unchanged, so no PKGREVISION bump done
txt2pdbdoc is a Unix-based Text-to-Doc file conversion program. (It
also converts Doc files to plain text.) A Doc file is a defacto
standard file format for text documents for PalmOS document reader
applications. The motivation for writing txt2pdbdoc was that such
software at the time was crufty and poorly documented. In contrast,
txt2pdbdoc is well written and well documented. Unix-style "man"
pages are included for the program, utility scripts, and the Doc file
format itself. The source code is also generously commented.
$Revision: 0.86 $ $Date: 2004/06/20 09:54:33 $
! Unicode/uni.c Unicode/uni2euc.h
Flat-table representation of uni2euc is now segmented to paged
tables so Visual C++ happy. Also pages w/ undefined characters
only are aggregated to shrink the table by more than 50%.
Thank you, Takahashi-san.
Message-Id: <006301c45664$a048cb20$0401a8c0@makoto>
0.85 2004/06/18 18:32:19
! Jcode.pm
getcode() is now less likely to fail to detect utf-8
Message-Id: <001901c45541$234adb00$0401a8c0@makoto>
2003/06/21 07:37:54
! Unicode/* Manifest
Unicode XS is completely rewritten from scratch. It is now
even faster than Encode (but less general, of course).
Instead of bsearch(), Unicode XS now uses flat lookup table
Which makes it not only faster but (hopefully) more portable.
Its only price is slightly larger size of the shared library
(360kb on FreeBSD 4-Stable)
Perl extension to deal with IMAP UTF7
IMAP mailbox names are encoded in a modified UTF7 when names contains
international characters outside of the printable ASCII range. The
modified UTF-7 encoding is defined in RFC2060 (section 5.1.3).
for Linux. Most (all?) Linux distros don't provide a libiconv
package, because GLIBC's iconv support is good enough.
At this time, using PKG_FAIL_REASON instead of PKG_SKIP_REASON
so we can catch examples of packages depending on this.
The builtin.mk was improved to help stop it from being used under
Linux.
This was okayed by Rene Hexel, the maintainer.
Todo: fix abiword build because it should not require this libiconv
package under Linux.
GLIBC supports GNU libiconv's API. So don't depend on pkgsrc's
libiconv for it. This is normal: most other Linux distros don't
provide a libiconv package.
This was discussed on tech-pkg list over past two months. It has
helped a few users under Linux where they had some problems with
conflicts with their working iconv() support provided with libc
and the libiconv package.
This was okayed by maintainer, Rene Hexel.
This still needs some improvement. Some packages still try to
force libiconv usage even though not needed.
breaks some things on Darwin) and fix audio/musicpd so it doesn't need
it (by faking out a configure check that gets confused by the
iconv_open -> libiconv_open renaming)
Changes since 2.23:
===================
2004-03-29 Gisle Aas <gisle@ActiveState.com>
Release 3.01
By compiling the extension with PERL_NO_GET_CONTEXT we can
make it slightly faster on a threaded perl. No change on a
regular perl. Patch provided by <beau@beaucox.com>.
Fixed missing ";" with assert. Patch provided by
Brendan O'Dea <bod@debian.org>.
Release 3.00
Drop the pure Perl implementations of the encoders and
decoders. They are bloat that hides real problems in
the XS implementations. I will re-release them separately
in the new MIME-Base64-Perl distribution.
The 'gcc -Wall' fix in 2.22 broke support for perl5.005,
as the isXDIGIT() macro is not available in that perl.
This problem has now been fixed.
