Upstream changes:
2.033 2017-07-06
- [RT #122371] Remove a couple of improperly-placed weaken statements
(reported by Phil Perry).
- [RT #122372] Fix weakening when a page is added to the end of a multiple
page document (reported by Phil Perry).
- Fix Bank Gothic core font (reported by Phil Perry).
2.032 2017-07-02
- PDF::API2 has many circular references, and the end() method doesn't clear
them all, so memory is leaked. This release uses Scalar::Util's weaken()
function to improve garbage collection. A significant number of circular
references have been weakened, though many likely still remain.
- [RT #120756] Eliminate a warning for an ambiguous call to CORE::open
(first reported by Abdelbaki Brahmia).
- $text->text_justified() and $text->text_fill_justified() now adjust the
space between words rather than stretching individual characters in order
to get the text to fit.
- [RT #120397] Indirect references and indirect objects can have comments
embedded in their whitespace, and their object number and generation may
be split across multiple lines, which may not all be buffered (reported by
SPROUT).
- [RT #120450] Fix PDF::API2->open($filename)->stringify() (reported by
SPROUT).
- Fix an off-by-one error when calculating text width while charspace is
non-zero.
- [RT #120048] Fix PDF::API2->synfont() (broken in 2.029, fixed by Vadim
Repin) and add basic testing.
Upstream changes:
2.031 2017-01-26
- Fix use of cache files when reading streams: temp files will now
be used any time a stream is larger than 16MB (by default).
Formerly, due to a bug, they would only be created when a 4kB
chunk of a stream increased to 16kB or more after being
decompressed.
- Numbers, booleans, and null values can now be read from object
streams.
- Update to [RT #113290]: Objects inside a large object stream are
now read without loading the entire object stream into memory.
- DEPRECATION: The low-level new_api methods have been deprecated
in favor of calling new directly. If your code uses
new_api($api2, ...), replace it with new($api2->{'pdf'}, ...).
- [RT #118352] Don't crash when adding an annotation to a page
that has an existing annotations array stored as an indirect
object (reported by Johan Vromans).
- [RT #118717] Die with an informative error if a file can't be
opened during open() or saveas() (reported by Johan Vromans).
Upstream changes:
2.030 2016-10-13
- Fix a font naming issue introduced while satisfying
Perl::Critic.
2.029 2016-10-10
- [RT #113293] Files with cross-reference streams weren't
correctly setting the max object number (report and
troubleshooting by Marco Pessotto).
- Handle TIFF images with strips that are wider than the image
(report and patch by Jeffrey Ratcliffe).
- [RT #98574] Increase test coverage of PDF::API2::Content (tests
by Phil Perry).
- A bunch of code cleanup and documentation updates by Paul
Cochrane.
- Add a missing prereq on Win32 systems (patch by Michiel Beijen).
- [RT #113514, #98552] Fix the dash() and renderingintent()
methods in ExtGState (reported by Vadim Repin and Phil Perry).
- Satisfy all Perl::Critic severity 5 policies.
- [RT #117940] Allow PNG, GIF, and PNM files to be opened from
filehandles in addition to filenames (patch by Johan Vromans).
- [RT #33970] Fail fast when a referenced file can't be opened
(requested by Barrie Slaymaker a mere 8.5 years ago).
- Add -simplex, -duplexfliplongedge and -duplexflipshortedge as
options to $pdf->preferences() (requested by Doug Poulin).
-------------------------------------
2.028 2016-06-08
- [RT #113290] Fix for reading objects inside an object stream in
a large PDF (patch by Marco Pessotto).
- Eliminate an infinite loop when reading a corrupt dictionary.
Upstream changes:
2.027 2016-03-11
- This release contains seven fixes for parsing PDFs. They mostly
affect files using cross-reference streams, which were first
supported in the previous release. Thanks to Marco Pessotto and
Stuart Henderson for their help identifying and troubleshooting
bugs.
- Added contrib/pdf-debug.pl to help track down issues related to
opening and parsing PDFs.
2.026 2016-02-24
- [RT #48683] Add support for PDFs with cross-reference streams
and object streams (patch by Don Huettl of Grant Street Group).
- [RT #107333] Accept an empty string as a valid Name, per PDF
spec 1.7 section 7.3.5 (patch by Mark Balitsky).
- [RT #98551] Rename aliases 2A, 4A, 2B, and 4B to 2A0, 4A0, 2B0,
and 4B0. The old names will continue to work, but are now
undocumented (patch by Phil Perry).
- Add -mils and -color as options for barcodes. Reset linedash so
that barcodes are always solid lines (patch by Erelen).
- [RT #98549] Rename meterlimit to miterlimit (patch by Phil
Perry).
- [RT #98534] Rename hspace to hscale (patch by Phil Perry).
- Fixed an infinite loop when RunLengthDecode is used for output.
- [RT #65582] Fix embedding of OpenType fonts (fix by Simon Cozens).
- [RT #67767] Allow an empty page to be imported into a PDF using
importPageIntoForm (reported by Antti Lankila).
- [RT #66341] Various fixes for ASCII85Decode and LZWDecode.
Problems found locating distfiles:
Package cabocha: missing distfile cabocha-0.68.tar.bz2
Package convertlit: missing distfile clit18src.zip
Package php-enchant: missing distfile php-enchant/enchant-1.1.0.tgz
Otherwise, existing SHA1 digests verified and found to be the same on
the machine holding the existing distfiles (morden). All existing
SHA1 digests retained for now as an audit trail.
Upstream changes:
2.025 2015-09-23
- The previous release included a patch that broke compatibility
with Perl 5.8.x. The minimum supported version is now Perl
5.8.5, and there is a .perl-version file to facilitate testing
using plenv.
2.024 2015-09-18
- [RT #104133] Include all bounding boxes when importing a page
(patch by Don Huettl of Grant Street Group).
- [RT #104133] When retrieving bounding boxes for a page, fall
back according to the defaults listed in the PDF spec (patch by
Don Huettl of Grant Street Group).
- Barcodes now take an optional -caption argument that will be
printed beneath the barcode. This can replace or be used in
conjunction with the text representation of the barcode
(requested by Gareth Tunley).
- [RT #105581] Calls to width() in BaseFont are now significantly
faster (patch by Dmitri Tikhonov).
{perl>=5.16.6,p5-ExtUtils-ParseXS>=3.15}:../../devel/p5-ExtUtils-ParseXS
since pkgsrc enforces the newest perl version anyway, so they
should always pick perl, but sometimes (pkg_add) don't due to the
design of the {,} syntax.
No effective change for the above reason.
Ok joerg
---------------------
2.023 2014-09-12
- The fix for [RT #69503] broke a previously-working case where a
page object could be passed (and was expected). The -firstpage
option now accepts a page object or a page number. Thanks to
Marco Pessotto for the bug report and test.
2.022 2014-07-04
- Added $pdf->version() get/set method. When opening an existing
PDF, the existing version number will now be retained.
- Renamed the following in PDF::API2:
- importpage to import_page
- openScalar to open_scalar
The old names are deprecated.
- [RT #69503] Fix the -firstpage option to $pdf->preferences() so
that it doesn't always lead to a crash (reported by Dietrich
Streifert).
- [RT #47974] Accept malformed xref subsections (with a warning)
that have extraneous spaces on the first line (reported by
Abhinav Kaushik).
- [RT #94505] For Code128 barcodes, the initial character set is
now optional (defaults to B or C depending on the content to be
encoded). The initial character set can be passed as a capital
letter, and the program will die if an invalid character set is
requested (requested by Andrea Nall).
- Interleaved 2 of 5 barcodes now prepend a zero when an odd
number of digits is specified, which is standard behavior.
Previously, a zero was appended instead.
- $page->rotate(0) now sets rotation to 0 degrees rather than
deleting an existing page rotation command. Page rotation is
inherited, so this is necessary to undo any inherited page
rotation.
- Fix: Attempts to use EAN-128 barcodes resulted in an error.
- Add a more informative error when text() is called without first
setting a font(), and when font() is called without including a
font size.
Do it for all packages that
* mention perl, or
* have a directory name starting with p5-*, or
* depend on a package starting with p5-
like last time, for 5.18, where this didn't lead to complaints.
Let me know if you have any this time.
Upstream changes:
2.021 2014-02-20
- Fixed numerous bugs in the string parsing code, including the
one reported in [RT #63918] by Frank Doepper.
- [RT #41049] Rewrote literal string parsing to prevent a stack
overflow due to an inefficient regex (reported by Sergei
Fetisov).
- [RT #91822] Fix compression of GIF images to ensure that output
codes don't exceed 12 bits (reported by Vadim Repin).
- The RunLengthDecode filter didn't actually work. Its code has
been rewritten and now passes basic encoding and decoding tests.
- Fix Code128 barcode switching from Code C to Code B in certain
cases (reported by Doru Petrescu).
a) refer 'perl' in their Makefile, or
b) have a directory name of p5-*, or
c) have any dependency on any p5-* package
Like last time, where this caused no complaints.
Upstream changes:
2.020 2013-01-20
- Give a more informative error message when a PDF file using a
cross-reference stream is encountered. The Known Issues section
of the documentation includes pointers on how to add support in
case someone else can get to this before I do.
- Text using TrueType fonts is now searchable again. In version
0.61, an undocumented -unicodemap option was added to nearly all
of the font resource methods, which only included a ToUnicode
CMap if it was set (and it's required in order for PDF readers
to be able to identify individual characters if BaseEncoding
isn't present, which it isn't for TrueType fonts).
I've left the -unicodemap option in place, and it's still
undocumented (except here), but it's now on by default. Call
$pdf->ttfont($fontfile, -unicodemap => 0) if you want to disable
it for performance or file size reasons.
- Add a note to the stringify method's documentation saying that
it's a destructive operation.
- The various filter types have their own classes in the
PDF::API2::Basic::PDF::Filter namespace now, rather than having
all of the packages in Filter.pm.
- To facilitate testing and ensure that identically-generated PDFs
have identical output in Perl 5.17.5 or later, PDF dictionary
keys are now sorted during output.
* Fix a typo in the require statement for code128 barcodes.
* Numerous modules are now only loaded when they're needed, which should help
speed up the load time and reduce the memory footprint.
* Test coverage is up to about 45% of the codebase.
Changes 2.018:
* The tests in 2.017 exposed a floating point issue where some computers give
more trailing zeros than others. It doesn't affect the output, but it does
break the tests on those computers, since the PDFs are slightly different.
This version modifies the float() method to remove trailing zeros, which
should fix the test breakage and give consistent PDFs.
Changes 2.017:
* The DejaVu fonts have been removed from the distribution, since they were
only used by one example script. If you need them for your project, you can
download them from http://dejavu-fonts.org
* Fix: Content->bogen() didn't behave as documented if $move was set (it
started drawing from the center of the circle rather than [x1, y1]).
* The undocumented nettfont method has been removed, along with the supporting
PDF::API2::Resource::Font::neTrueType module.
* The undocumented save_xml method has been removed, along with the supporting
functions in PDF::API2::Basic::*.
* The undocumented textstate2 method has been removed from PDF::API2::Content.
* Fix create_egs call in Lite.pm
* Fix string interpolation in pdf-merge.pl
* Fix an error in the synopsis
* Fix rename of .cmap files so that they actually work now.
* The test suite has grown to nearly 200 tests, covering about 40% of the
codebase. There's still plenty of room for improvement.
to trigger/signal a rebuild for the transition 5.10.1 -> 5.12.1.
The list of packages is computed by finding all packages which end
up having either of PERL5_USE_PACKLIST, BUILDLINK_API_DEPENDS.perl,
or PERL5_PACKLIST defined in their make setup (tested via
"make show-vars VARNAMES=..."), minus the packages updated after
the perl package update.
sno@ was right after all, obache@ kindly asked and he@ led the
way. Thanks!
- removed packages p5-IO-Compress-Base, p5-IO-Compress-Zlib,
p5-IO-Compress-Bzip2 and p5-Compress-Zlib because they are
merged into p5-IO-Compress
- Updated dependend packages to depend on p5-IO-Compress
and bump PKGREVISION
Upstream changes:
2.017 30 March 2009
* Merged IO-Compress-Base, IO-Compress-Bzip2, IO-Compress-Zlib &
Compress-Zlib into IO-Compress.
* The interface to Compress-Raw-Zlib now uses the new LimitOutput
feature. This will make all of the zlib-related IO-Compress modules
less greedy in their memory consumption.
* Removed MAN3PODS from Makefile.PL
* A few changes to get the test harness to work on VMS courtesy of
Craig. A. Berry.
* IO::Compress::Base & IO::Uncompress::Base
Downgraded some croaks in the constructors to just set $! (by letting
the code attempt to open a file and fail).
This makes the behavior more consistent to a standard open.
[RT #42657]
* IO::Uncompress::Base
Doing a seek with MultiStream could drop some of the uncompressed
data. Fixed.
* IO::Compress::Zip
- Fixed problem with the uncompressed & uncompressed fields when
zip64 is enabled. They were set to 0x0000FFFF instead of
0xFFFFFFFF. Also the ZIP64 extra field was 4 bytes short.
Problem spotted by Dino Chiesa.
* IO::Uncompress::Unzip
- use POSIX::mktime instead of Time::Local::timelocal to convert
the zip DOS time field into Unix time.
* Compress::Zlib
- Documented Compress::Zlib::zlib_version()
Pkgsrc changes:
o Tweak MASTER_SITES so that the distfile is found
Upstream changes:
No change log available. From looking at the diff and the RT database:
o Improve forward compatibility with perl 5.10.0 maintenance releases
(RT #39875?)
o fixed [rt.cpan.org #40648] Unicode text prints text on top of text before it
to trigger/signal a rebuild for the transition 5.8.8 -> 5.10.0.
The list of packages is computed by finding all packages which end
up having either of PERL5_USE_PACKLIST, BUILDLINK_API_DEPENDS.perl,
or PERL5_PACKLIST defined in their make setup (tested via
"make show-vars VARNAMES=...").
The upstream package does not have a change log, and there
is already an open RT ticket for this problem as well.
A diff doesn't reveal anything which appears to be earth-shattering.
Pkgsrc changes:
- List version independent HOMEPAGE.
- The package supports installation to DESTDIR.
- The module needs at least Compress::Zlib 1.0.
Changes since version 0.65:
===========================
not provided.
developer is officially maintaining the package.
The rationale for changing this from "tech-pkg" to "pkgsrc-users" is
that it implies that any user can try to maintain the package (by
submitting patches to the mailing list). Since the folks most likely
to care about the package are the folks that want to use it or are
already using it, this would leverage the energy of users who aren't
developers.
module directory has changed (eg. "darwin-2level" vs.
"darwin-thread-multi-2level").
binary packages of perl modules need to be distinguishable between
being built against threaded perl and unthreaded perl, so bump the
PKGREVISION of all perl module packages and introduce
BUILDLINK_RECOMMENDED for perl as perl>=5.8.5nb5 so the correct
dependencies are registered and the binary packages are distinct.
addresses PR pkg/28619 from H. Todd Fujinaka.
