A substantial revision of the build system and GhostPDL directory structure, meaning the GhostPCL and GhostXPS "products" are now built by the Ghostscript build system "proper" rather than having their own builds which called the Ghostscript build system for certain parts of the builds.
Ghostscript-only users should not be affected by this change.
For users of the PCL and XPS interpreters (collectively downloaded as "GhostPDL") there is a new, "flatter" directory structure and only one configure and one make invocation to build all the products. There are also some executable default name changes to bring the PCL and XPS binaries in line with the Ghostscript ones. All executables now end up in the "bin" directory (or build specific variant thereof, e.g. "debugbin"). The PCL executable, by default, is now "gpcl6" on Unix like systems and "gpcl6win??.exe" (where "??" is either "32" or "64"). The XPS executable remains "gxps" on Unix-like systems, but similarly becomes "gxpswin??.exe" on Windows.
Those using scripts or other automated processes will need to take steps to cope with these changes.
A new method of internally inserting devices into the device chain has been developed, named "device subclassing". This allows suitably written devices to be more easily and consistently as "filter" devices.
The first fruit of this is a new implementation of the "-dFirstPage"/"-dLastPage" feature which functions a device filter in the Ghostscript graphics library, meaning it works consistently with all input languages.
The "ghostpdl.inf" file, provided to simplify installing a Postscript printer configure to produce output to suit Ghostscript's "pdfwrite" device is now available as a digitally signed version: "ghostpdl.cat".
Plus the usual round of bug fixes, compatibility changes, and incremental improvements.
* Update HOMEPAGE and MASTER_SITES.
Changelog:
pdf2djvu (0.8.2) unstable; urgency=low
[ Jakub Wilk ]
* Fix file corruption for some documents with outline.
https://bitbucket.org/jwilk/pdf2djvu/issues/110
Regression introduced in 0.7.20.
* Improve thread-safety:
+ Make GraphicsMagick initialization thread-safe.
+ Rewrite the subprocess handling code to be thread-safe.
Remove PStreams support.
* Fix silencing djvuextract's stderr.
* Use iconv() for converting from locale encoding to UTF-8.
* Document that gettext is needed for internationalization.
* Improve error handling.
* Improve the test suite.
* Update the Polish translation.
[ Américo Monteiro ]
* Update the Portuguese translation.
[ Chris Leick ]
* Update the German translation.
-- Jakub Wilk <jwilk@jwilk.net> Tue, 08 Sep 2015 20:43:24 +0200
pdf2djvu (0.8.1) unstable; urgency=low
* Fix “--fg-colors=N” when GraphicsMagick was compiled with
QuantumDepth > 8.
* Report GraphicsMagick's QuantumDepth in “--version”.
* Build with “-Werror=narrowing”.
* Document what software is required to run the tests.
* Windows: fix loading Poppler data files.
* Improve the test suite.
-- Jakub Wilk <jwilk@jwilk.net> Thu, 13 Aug 2015 17:48:55 +0200
pdf2djvu (0.8) unstable; urgency=low
[ Juan Francisco Cantero Hurtado ]
* Don't hardcode the Python interpreter path in script shebangs; use
“#!/usr/bin/env python” instead.
[ Jakub Wilk ]
* Don't silently ignore --loss-level when used without --monochrome.
https://bitbucket.org/jwilk/pdf2djvu/issue/86
* Add sanity check that output file is distinct from all input files.
https://bitbucket.org/jwilk/pdf2djvu/issue/98
* Apply NFKC normalization on characters from the Alphabetic Presentation
Forms block (U+FB00–U+FB4F), even when --no-nfkc were specified.
This is what Poppler (≥ 0.19) does by default.
Thanks to Dmitry Shachnev for the bug report.
https://bitbucket.org/jwilk/pdf2djvu/issue/90https://bugs.freedesktop.org/show_bug.cgi?id=7002
* Fail early if duplicate page numbers were specified in -p/--pages.
https://bitbucket.org/jwilk/pdf2djvu/issue/73
* Fail early on duplicate page identifiers and duplicate page titles.
* Make --help and --version return with exit code 0.
* Windows: build against GraphicsMagick.
https://bitbucket.org/jwilk/pdf2djvu/issue/75
* Improve the configure script:
+ Check if Poppler xpdf headers exist.
+ Add option for disabling PStreams (--without-pstreams).
* Change the order libraries on linker command line.
Move libraries provided by the LIBS configure variable to the end.
* Use the reentrant S-expression output API when it's available
(DjVuLibre ≥ 3.5.25).
* Improve error handling.
* Improve the test suite.
* Improve documentation:
+ Simplify documentation about build dependencies.
+ Document that pkg-config is required for building.
+ Update poppler-data download URL.
* Update the Polish translation.
[ Américo Monteiro ]
* Update the Portuguese translation.
[ Chris Leick ]
* Update the German translation.
[ Kyrill Detinov ]
* Update the Russian translation.
[ Jean-Paul Guillonneau ]
* Update the French translation of the manual page.
-- Jakub Wilk <jwilk@jwilk.net> Fri, 07 Aug 2015 14:42:08 +0200
pdf2djvu (0.7.21) unstable; urgency=low
* Fix insecure use of /tmp when executing c44(1).
https://bitbucket.org/jwilk/pdf2djvu/issue/103
-- Jakub Wilk <jwilk@jwilk.net> Sat, 09 May 2015 23:21:40 +0200
pdf2djvu (0.7.20) unstable; urgency=low
[ Jakub Wilk ]
* Don't use djvused for setting document outline.
It was too slow for very long documents.
https://bitbucket.org/jwilk/pdf2djvu/issue/93
* Test suite: automatically skip tests that require features that were not
enabled at build time.
https://bitbucket.org/jwilk/pdf2djvu/issue/55
* Improve reporting of versions of the libraries in --version and XMP
metadata.
* Improve Windows support:
+ Correctly determine path to DjVuLibre binaries.
https://bitbucket.org/jwilk/pdf2djvu/issue/85
+ Fix path splitting is East-Asian locales.
https://bitbucket.org/jwilk/pdf2djvu/issue/100
+ Fix stdout redirection.
https://bitbucket.org/jwilk/pdf2djvu/issue/101
* Update the homepage URL and the bug tracker URLs, in documentation and
code.
http://google-opensource.blogspot.com/2015/03/farewell-to-google-code.html
[ Américo Monteiro ]
* Update the Portuguese translation.
[ Kyrill Detinov ]
* Update the Russian translation.
[ Chris Leick ]
* Update the German translation.
[ Serhij Dubyk ]
* Update the Ukrainian translation.
-- Jakub Wilk <jwilk@jwilk.net> Fri, 01 May 2015 14:21:16 +0200
pdf2djvu (0.7.19) unstable; urgency=low
* Fix build failure with Poppler ≥ 0.31.
Thanks to David Evans for the bug report.
https://bitbucket.org/jwilk/pdf2djvu/issue/99
-- Jakub Wilk <jwilk@jwilk.net> Fri, 06 Feb 2015 20:00:38 +0100
pdf2djvu (0.7.18) unstable; urgency=low
[ Jakub Wilk ]
* Make the xml2c script more portable.
* In C++11 mode, use “decltype” instead of “typeof”.
* Fix build failure with libc++ in C++98 mode.
Thanks to Angelo Graziosi for the bug report.
https://bitbucket.org/jwilk/pdf2djvu/issue/88
* Use the standard AC_OPENMP macro in the configure script (instead of
3rd-party AX_OPENMP).
* Use HTTPS URLs when they are available, in documentation and code.
* Fix capitalization error in the German manual page.
Thanks to Chris Leick for help.
* Drop support for Poppler < 0.11.1.
* Improve error handling.
[ Américo Monteiro ]
* Add Portuguese translation.
https://bugs.debian.org/758262
[ Majid Redouane, Jean-Paul Guillonneau ]
* Add French translation of the manual page.
-- Jakub Wilk <jwilk@jwilk.net> Thu, 18 Dec 2014 13:25:10 +0100
pdf2djvu (0.7.17) unstable; urgency=low
* Fix build failure with Poppler ≥ 0.23.
-- Jakub Wilk <jwilk@jwilk.net> Thu, 30 May 2013 22:20:31 +0200
pdf2djvu (0.7.16) unstable; urgency=low
* Fix build failure with Poppler ≥ 0.21.1.
-- Jakub Wilk <jwilk@jwilk.net> Wed, 14 Nov 2012 20:44:57 +0100
pdf2djvu (0.7.15) unstable; urgency=low
* Fix build failure with Poppler ≥ 0.21.
-- Jakub Wilk <jwilk@jwilk.net> Sat, 06 Oct 2012 12:32:09 +0200
Changelog:
2015-05-24 Jay Berkenbilt <ejb@ql.org>
* 5.1.3: release
* Bug fix: fix-qdf was not handling object streams with more than
255 objects in them.
* Handle Microsoft crypt provider initialization properly for case
where no keys have been previously created, such as in a fresh
Windows installation.
* Include time.h in QUtil.hh for time_t
PrawnPDF 2.0.2 -- 2015-07-15
Links in repeaters/stamps are now clickable
Previously, url links were not clickable when rendered within a stamp. The proper annotation references are now added to the page object that the stamp is called, thereby generating a clickable link in the pdf.
Because repeaters are built upon stamps, this fix should also solve issues with links inside of repeaters.
(Jesse Doyle, #801, #498)
* Add spec for trailer initialization
* Update document state spec description
* Remove securerandom
* Fixed tests for document state trailer
* Remove default_trailer & leave it as an empty hash
* Add document ID to trailer for PDF-X spec compat
Remove unnecessary autoconf dependency.
Bug report for building with BSD make was fixed, so remove patch.
CHANGES IN V1.0.76
- cups-browsed: Make build also working with BSD make (Bug
#1310).
- cups-browsed, implicitclass: Let the load-balancing queue up
the jobs in the local (generated) queue until a free (idle,
enabled, and accepting jobs) remote queue is found (check
every 5 sec if no free queue available). This gives a more
even distribution of the work amongst the servers and
protects against the case that a bunch of jobs gets
inaccessible or lost if one of the servers fails.
- cups-browsed: Let the load-balancing also check whether the
destination queue is actually accepting jobs.
- cups-browsed: If a generated queue is not removed on
shutdown of cups-browsed due to remaining jobs in it,
re-enable it in the next cups-browsed session even if it was
disabled by something other than cups-browsed.
- implicitclass: Clean up debug and error messages.
- implicitclass: Fix exit codes for immediate retry of jobs.
- cups-browsed: Make absence of CUPS notifications via D-Bus
non-fatal and fall back to the old behavior of cups-browsed
(fail-over instead of load balancing for equally-named
remote queues, do not remove generated queue if it is
default instead of default printer management). This solves
problems of CUPS and/or cups-browsed built without D-Bus
support or absence of D-Bus on the system (Bug #1316).
- cups-browsed: Do not use g_warning() function, sneaked in by
copy and paste of CUPS subscription functions.
CHANGES IN V1.0.75
- texttopdf: Really support BoldItalic (original texttops just maps
BoldItalic to Bold).
- texttopdf: Fixed segfault when outputting BoldItalic (Bug #1314).
Gutenprint 5.2.10 no longer supports Macintosh OS X 10.4 or earlier.
[See NEWS for the list of printers; ~1000 lines were omitted from this
commit message.]
1) Added a unified CUPS backend to support selected dye sublimation
printers.
2) New printers supported in this release:
[many]
3) The following Canon printer models were not added, as they cannot
currently be supported, even experimentally, owing to an as-yet
unknown command format or other issues:
[several]
4) Corrected page dimensions for borderless printing with Canon
inkjet printers.
5) Added duplex support for the EPSON WorkForce 630, 635, and 645,
and NX635
6) Expanded printable area at the bottom of the page for the Epson
Stylus Pro 3800 and 3880.
7) Added IEEE1284 device IDs for the following printers.
[many]
8) Added Catalan, Turkish, and Vietnamese translations.
CHANGES.txt - 2.1.0 - 2015-08-31
--------------------------------
CHANGES IN CUPS V2.1.0
- Fixed more scheduler crash bugs in the new logging code (STR #4687,
STR #4690)
- The scheduler did not use the ConfigFilePerm setting when copying PPD
files or interface scripts attached to a request (STR #4703)
- Now support new Chinese locale IDs and their correct fallback locales
(<rdar://problem/22086642>, <rdar://problem/22130168>)
- "make check" incorrectly reported an expectation of 18 warning
messages when 8 were expected (STR #4684)
- The new PDF file type rule did not work (STR #4692)
- The scheduler did not update the jobs.cache file when job files were
expired (STR #4706)
- Fixed some configure script issues (STR #4694, STR #4695, STR #4698)
- Documentation updates (STR #4691, STR #4693)
CHANGES IN CUPS V2.1rc1
- Added support for 3D printers (basic types only, no built-in filters)
based on PWG white paper.
- Fixed bugs in the new journald support (STR #4655, STR #4658,
STR #4661)
- Fixed domain socket support on Linux (STR #4679)
- Fixed signal handlers in the dnssd and usb backends (STR #4671)
- <Limit All> in <Policy> sections now applies to all operations when
used by itself (STR #4659)
- Configure script changes for systemd support (STR #4669)
- Updated autoconf sources to use newer form of AC_INIT (STR #4664)
CHANGES IN CUPS V2.1b1
- Improved speed of ppdMarkDefaults for complex/large PPDs
(<rdar://problem/15146999>)
- The IPP backend now stops sending print data if the printer indicates
the job has been aborted or canceled (<rdar://problem/17837631>)
- The IPP backend now sends the job-pages-per-set attribute when
printing multiple copy jobs with finishings
(<rdar://problem/16792757>)
- The IPP backend now updates the cupsMandatory values when the printer
configuration changes (<rdar://problem/18126570>)
- No longer install banner files since third-party banner filters now
supply their own (STR #4518)
- Added support for EXPECT-ALL directive in ipptool test files
(STR #4469)
- Added support for WITH-VALUE-FROM predicate in ipptool test files
(STR #4470)
- The scheduler no longer listens on the loopback interface unless the
web interface or printer sharing are enabled
(<rdar://problem/9136448>)
- Added a PPD generator for IPP Everywhere printers (STR #4258)
- Now install "default" versions of more configuration files
(<rdar://problem/19024491>)
- The cupstestppd program did not handle "maxsize(nnn)" entries in
cupsFilter/cupsFilter2 values (<rdar://problem/18974858>)
- The scheduler now checks the return value of rename() calls
(STR #4589)
- The scheduler now validates ErrorPolicy values in config files
(STR #4591)
- Long cookies caused the web interface to stop working (STR #4619)
- Added SSLOptions values to allow Diffie-Hellman key exchange and
disable TLS/1.0 support.
- Updated the scheduler to support more IPP Everywhere attributes
(STR #4630)
- The scheduler now supports advanced ASL and journald logging when
"syslog" output is configured (STR #4474)
- The scheduler now supports logging to stderr when running in the
foreground (STR #4505)
Add tex-toptesi{,-doc} 5.86f Bundle for typesetting multilanguage theses
Add tex-stringstrings{,-doc} 1.23 String manipulation for cosmetic and
programming application
Add tex-probsoln{,-doc} 3.04 Generate problem sheets and their solution sheets
Add tex-flowfram{,-doc} 1.17 Create text frames for posters, brochures or
magazines
Add tex-europasscv{,-doc} 2015 Unofficial class for the new version of the
Europass curriculum vitae
Add tex-envlab{,-doc} 1.2 Addresses on envelopes or mailing labels
Add tex-newlfm{,-doc} 9.4 Write letters, facsimiles, and memos
Add tex-cuisine{,-doc} 0.7 Typeset recipes in LaTeX
From yhardy in pkgsrc-wip. Cleanup by me.
CHANGES IN V1.0.74
- cups-browsed: Added NULL check when getting the notification of
a printer starting to process a job and checking whether this
printer is created by cups-browsed with the implicitclass:
backend (Ubuntu bug #1488524).
- backends: Include unistd.h and fcntl.h in backend-private.h for
all platforms, not only Linux, so that the backends build also
on non-Linux platforms (Bug #1308).
- cups-browsed: Do not schedule printers for update when they are
already marked as disappeared.
- cups-browsed: Added sanity checks when saving the default printer
selection in a file.
- cups-browsed, implicitclass: If remote queues disappear or
cups-browsed shuts down and a cups-browsed-generated queue still
has jobs and needs to be kept therefore, disable it and re-enable
it when cups-browsed starts again and/or the remote queue(s) re-
appear(s). This avoids repeated retries of the jobs while the
remote server is not available, causing unneeded system load and
battery consumption.
- cups-browsed: After polling info from remote CUPS servers to find
the best destination job, set the default CUPS server back to local.
These changes:
- fix paths for /etc and /var after those of Python installer files
- while there, fix paths in additional files
- locate libraries on systems without ldconfig(8)
This is particularly relevant for some of the tools installed, like
hp-check (which still doesn't complete). With these changes, I can print
again with my M276n printer.
CHANGES IN V1.0.73
- cups-browsed: Added missing
utils/org.cups.cupsd.Notifier.xml file.
CHANGES IN V1.0.72
- cups-browsed, implicitclass: Added load balancing
functionality. If there are several remote CUPS printers
with the same name, they for locally a cluster represented
by a print queue with this name. This printer prints through
a special backend (implicitclass) which makes cups-browsed
find the best destination remote queue (fewest jobs,
enabled) for this job.
- cups-browsed: Added protection against accidental deletion of
print queues generated by cups-browsed. These queues now get
automatically re-created.
- cups-browsed: Added LDAP support. Appropriately configured via
cups-browsed.conf remote printers made available via LDAP will be
looked up and local queues pointing to them created. Thanks to
Raphael Geissert (atomo64 at gmail dot com) for contributing this
patch (Debian bug #795185).
- cups-browsed: Introduced new handling for the default
printer using cache files. So we do not need to keep an auto-generated
queue because it is set as default printer. If the auto-generated
queue disappears, the old local printer is set as default again and
when it re-appears it returns to be the default printer.
- cups-browsed: Added infrastructure for subscribing to CUPS
notifications for things like improved default printer handling,
load balancing, ...
- foomatic-rip: Prevent crash when supplying "media" option with empty
value ("media=", Ubuntu bug #1479871).
- pdftoopvp: Adaptations to API changes on Poppler 0.34.0, note that
this patch disables color management in this filter. Thanks to
Vincent le Garrec and Andreas K. Huettel for the patch (Bug #1301,
Gentoo bug #554782).
- libcupsfilters, bannertopdf, foomatic-rip, gstoraster, pdftoijs,
sys5ippprinter, pdftoopvp, pdftops, pdftoraster, rastertoescpx,
urftopdf, texttopdf: Miscellaneous fixes for build compatibility with
different platforms, like config.h as very first include and so on.
Thanks to Richard Palo for the patch (Bug #1264).
- texttopdf: Request the generic 'monospace' font alias from fontconfig
instead of the hard-coded FreeMono. Thanks to Fabian Greffrath
(fabian at debian dot org) for the patch (Debian bug #788048).
Patch 1.25.1, 2015-07-20
- Fix bug when parsing inline images. Occurred when merging
certain pages with inline images
- Fixed type error when creating outlines by utilizing the
isString() test
Version 1.25, 2015-07-07
------------------------
BUGFIXES:
- Added Python 3 algorithm for ASCII85Decode. Fixes issue when
reading reportlab-generated files with Py 3 (jerickbixly)
- Recognize more escape sequence which would otherwise throw an
exception (manuelzs, robertsoakes)
- Fixed overflow error in generic.py. Occurred
when reading a too-large int in Python 2 (by Raja Jamwal)
- Allow access to files which were encrypted with an empty
password. Previously threw a "File has not been decrypted"
exception (Elena Williams)
- Do not attempt to decode an empty data stream. Previously
would cause an error in decode algorithms (vladir)
- Fixed some type issues specific to Py 2 or Py 3
- Fix issue when stream data begins with whitespace (soloma83)
- Recognize abbreviated filter names (AlmightyOatmeal and
Matthew Weiss)
- Copy decryption key from PdfFileReader to PdfFileMerger.
Allows usage of PdfFileMerger with encrypted files (twolfson)
- Fixed bug which occurred when a NameObject is present at end
of a file stream. Threw a "Stream has ended unexpectedly"
exception (speedplane)
FEATURES:
- Initial work on a test suite; to be expanded in future.
Tests and Resources directory added, README updated (robertsoakes)
- Added document cloning methods to PdfFileWriter:
appendPagesFromReader, cloneReaderDocumentRoot, and
cloneDocumentFromReader. See official documentation (robertsoakes)
- Added method for writing to form fields: updatePageFormFieldValues.
This will be enhanced in the future. See official documentation
(robertsoakes)
- New addAttachment method. See documentation. Support for adding
and extracting embedded files to be enhanced in the future
(moshekaplan)
- Added methods to get page number of given PageObject or
Destination: getPageNumber and getDestinationPageNumber.
See documentation (mozbugbox)
OTHER ENHANCEMENTS:
- Enhanced type handling (Brent Amrhein)
- Enhanced exception handling in NameObject (sbywater)
- Enhanced extractText method output (peircej)
- Better exception handling
- Enhanced regex usage in NameObject class (speedplane)
Version 1.24, 2014-12-31
------------------------
- Bugfixes for reading files in Python 3 (by Anthony Tuininga and
pqqp)
- Appropriate errors are now raised instead of infinite loops (by
naure and Cyrus Vafadari)
- Bugfix for parsing number tokens with leading spaces (by Maxim
Kamenkov)
- Don't crash on bad /Outlines reference (by eshellman)
- Conform tabs/spaces and blank lines to PEP 8 standards
- Utilize the readUntilRegex method when reading Number Objects
(by Brendan Jurd)
- More bugfixes for Python 3 and clearer exception handling
- Fixed encoding issue in merger (with eshellman)
- Created separate folder for scripts
ok wiz@.
pkgsrc changes:
o Added LICENSE
Changes:
2014-11-14 Till Kamppeter <till.kamppeter@gmail.com>
* foomatic-ppdfile.in: Foomatic doesn't provide some offered PPD
files. Thanks to Marek Kasik for the patch (bug #1238).
2014-08-22 Till Kamppeter <till.kamppeter@gmail.com>
* foomatic-ppd-to-xml.in: Let missing XML files be added when to a
PPD with already existing XML files new "*Product:" lines get
added.
2014-03-20 Till Kamppeter <till.kamppeter@gmail.com>
* Tagged branch for release 4.0.11.
* README, USAGE, configure.ac: Updated for release 4.0.11.
* lib/Foomatic/DB.pm: Do not interpret option default values set to
"0" in PPD files as no default setting defined. Thanks to Deng
Pang from Ricoh (DengPang at rst dot ricoh dot com) for the report.
2014-02-18 Till Kamppeter <till.kamppeter@gmail.com>
* Tagged branch for release 4.0.10.
* README, USAGE, configure.ac: Updated for release 4.0.10.
2014-01-02 Till Kamppeter <till.kamppeter@gmail.com>
* foomatic-addpjloptions.in: Make foomatic-addpjloptions work with
the system's Foomatic database, too.
no longer needed/required.)
Updated to 3.15.7
Release Notes:
HPLIP 3.15.7 - This release has the following changes:
Added Support for the Following New Printers:
- HP Deskjet 1110 Printer
- HP Deskjet 1111 Printer
- HP Deskjet 1112 Printer
- HP DeskJet 2131 All-in-One Printer
- HP Deskjet Ink Advantage 1110 Printer
- HP Deskjet Ink Advantage 1115 Printer
- HP Deskjet Ink Advantage 1118 Printer
- HP DeskJet Ink Advantage 2130 All-in_One Printer
- HP DeskJet Ink Advantage 2135 All-in_One Printer
- HP DeskJet Ink Advantage 2138 All-in_One Printer
- HP DeskJet Ink Advantage 3630 All-in-One Printer
- HP DeskJet Ink Advantage 3635 All-in-One Printer
- HP DeskJet Ink Advantage 3636 All-in-One Printer
- HP DeskJet Ink Advantage 3638 All-in-One Printer
- HP DeskJet Ink Advantage 3830 All-in-One Printer
- HP DeskJet Ink Advantage 3835 All-in-One Printer
- HP OfficeJet 3830 All-in-One Printer
- HP OfficeJet 3832 All-in-One Printer
- HP OfficeJet 3834 All-in-One Printer
- HP Officejet 7510 Wide Format All-in-One Printer
Added support for the following new Distro's:
- Debian 8.1
- RHEL 7.1
Issues fixed:
- Removed the duplicate collate option from HP Device Manager
- Traceback error occurs when trying to configure a queue of same name
but with different IO type
- Traceback error occurs when trying to add unicode file from Device
Manager on RHEL5
Launchpad fixes:
1464013 - HPLIP installer prompts to set up printers, even when printers
are already set up
1432516 - Insecure binary driver verification (CVE-2015-0839)
WeasyPrint is a visual rendering engine for HTML and CSS that can
export to PDF. It aims to support web standards for printing.
WeasyPrint is free software made available under a BSD license.
It is based on various libraries but not on a full rendering engine
like WebKit or Gecko. The CSS layout engine is written in Python,
designed for pagination, and meant to be easy to hack on.
