Alex Deucher (7):
radeon: enable hawaii accel conditionally (v3)
radeon: fix warnings when building against older xservers
radeon/kms: add new CIK pci ids
radeon/kms: add new SI pci ids
radeon/kms: fix mullins pci id
radeon: require libdrm_radeon 2.4.58 for latest SI/CI pci ids
radeon: bump version for release
Andreas Boll (4):
radeon: drop redundant radeon_drm.h includes
radeon: move RADEON_TILING_{MASK, LINEAR} from radeon_drm.h to radeon.h
radeon: drop radeon_drm.h
radeon: remove definitions already present in radeon_drm.h
Hans de Goede (1):
configure: Include xorg-server.h before glamor.h
Maarten Lankhorst (1):
bump version post release
Michel Dänzer (4):
Handle CRTC DPMS from output DPMS hooks
glamor: Set environment variable RADEON_THREAD=0
Add Emacs .dir-local.el file
Revert "glamor: Set environment variable RADEON_THREAD=0"
Thomas DeRensis (2):
radeon: fix a leak in radeon_vbo_get_bo()
radeon: fix build warnings regarding const qualifier
This is a regularly-scheduled bugfix release.
commit: catch changed exec bit on files from p1 (issue4382)
diff: document the nobinary option
dispatch: check shell alias again after loading extensions (issue4355)
dispatch: make "_checkshellalias" reusable regardless of adding aliases
help: fix typo in log examples
help: mention mode in hg log --removed help (issue4381)
hgweb: fail if an invalid command was supplied in url path (issue4071)
hgweb: refresh hgweb.repo on phase change (issue4061)
histedit: abort gracefully on --continue/--abort with no state
keepalive: fix how md5 is used
mq: examine "pushable" of already applied patch correctly
mq: pop correct patches when changing pushable-ness of already applied ones
mq: report correct numbers for changing "number of guarded, applied patches"
mq: use "mq.applied[i].name" instead of "mq.appliedname(i)" for safety
revset: add an optimised baseset.contains (issue4371)
templater: fix precedence of --style and --template options
* Moved ResultIterator/PageIterator to ccmain.
* Added Right-to-left/Bidi capability in the output iterators for Hebrew/Arabic.
* Added paragraph detection in layout analysis/post OCR.
* Fixed inconsistent xheight during training and over-chopping.
* Added simultaneous multi-language capability.
* Refactored top-level word recognition module.
* Added experimental equation detector.
* Improved handling of resolution from input images.
* Blamer module added for error analysis.
* Cleaned up externally used namespace by removing includes from baseapi.h.
* Removed dead memory mangagement code.
* Tidied up constraints on control parameters.
* Added support for ShapeTable in classifier and training.
* Refactored class pruner.
* Fixed training leaks and randomness.
* Major improvements to layout analysis for better image detection, diacritic detection, better textline finding, better tabstop finding.
* Improved line detection and removal.
* Added fixed pitch chopper for CJK.
* Added UNICHARSET to WERD_CHOICE to make mult-language handling easier.
* Fixed problems with internally scaled images.
* Added page and bbox to string in tr files to identify source of training data better.
* Fixes to Hindi Shiroreka splitter.
* Added word bigram correction.
* Reduced stack memory consumption and eliminated some ugly typedefs.
* Added new uniform classifier API.
* Added new training error counter.
* Fixed endian bug in dawg reader.
* Many other fixes, including the way in which the chopper finds chops and messes with the outline while it does so.
Upstream changes:
Version 1.23, 2014-08-11
------------------------
- Documentation now available at http://pythonhosted.org//PyPDF2
- Bugfix in pagerange.py for when __init__.__doc__ has no value (by
Vladir Cruz)
- Fix typos in OutlinesObject().add() (by shilluc)
- Re-added a missing return statement in a utils.py method
- Corrected viewing mode names (by Jason Scheirer)
- New PdfFileWriter method: addJS() (by vfigueiro)
- New bookmark features: color, boldness, italics, and page fit
(by Joshua Arnott)
- New PdfFileReader method: getFields(). Used to extract field
information from PDFs with interactive forms. See documentation
for details
- Converted README file to markdown format (by Stephen Bussard)
- Several improvements to overall performance and efficiency
(by mozbugbox)
- Fixed a bug where geospatial information was not scaling along with
its page
- Fixed a type issue and a Python 3 issue in the decryption algorithms
(with Francisco Vieira and koba-ninkigumi)
- Fixed a bug causing an infinite loop in the ASCII 85 decoding
algorithm (by madmaardigan)
- Annotations (links, comment windows, etc.) are now preserved when
pages are merged together
- Used the Destination class in addLink() and addBookmark() so that
the page fit option could be properly customized
Version 1.22, 2014-05-29
------------------------
- Added .DS_Store to .gitignore (for Mac users) (by Steve Witham)
- Removed __init__() implementation in NameObject (by Steve Witham)
- Fixed bug (inf. loop) when merging pages in Python 3 (by commx)
- Corrected error when calculating height in scaleTo()
- Removed unnecessary code from DictionaryObject (by Georges Dubus)
- Fixed bug where an exception was thrown upon reading a NULL string
(by speedplane)
- Allow string literals (non-unicode strings in Python 2) to be passed
to PdfFileReader
- Allow ConvertFunctionsToVirtualList to be indexed with slices and
longs (in Python 2) (by Matt Gilson)
- Major improvements and bugfixes to addLink() method (see documentation
in source code) (by Henry Keiter)
- General code clean-up and improvements (with Steve Witham and Henry Keiter)
- Fixed bug that caused crash when comments are present at end of
dictionary
Version 1.21, 2014-04-21
------------------------
- Fix for when /Type isn't present in the Pages dictionary (by Rob1080)
- More tolerance for extra whitespace in Indirect Objects
- Improved Exception handling
- Fixed error in getHeight() method (by Simon Kaempflein)
- implement use of utils.string_type to resolve Py2-3 compatibility issues
- Prevent exception for multiple definitions in a dictionary (with carlosfunk)
(only when strict = False)
- Fixed errors when parsing a slice using pdfcat on command line (by
Steve Witham)
- Tolerance for EOF markers within 1024 bytes of the actual end of the
file (with David Wolever)
- Added overwriteWarnings parameter to PdfFileReader constructor, if False
PyPDF2 will NOT overwrite methods from Python's warnings.py module with
a custom implementation.
- Fix NumberObject and NameObject constructors for compatibility with PyPy
(R▒diger Jungbeck, Xavier Dupr▒, shezadkhan137, Steven Witham)
- Utilize utils.Str in pdf.py and pagerange.py to resolve type issues (by
egbutter)
- Improvements in implementing StringIO for Python 2 and BytesIO for
Python 3 (by Xavier Dupr▒)
- Added /x00 to Whitespaces, defined utils.WHITESPACES to clarify code (by
Maxim Kamenkov)
- Bugfix for merging 3 or more resources with the same name (by lucky-user)
- Improvements to Xref parsing algorithm (by speedplane)
Version 1.20, 2014-01-27
------------------------
- Official Python 3+ support (with contributions from TWAC and cgammans)
Support for Python versions 2.6 and 2.7 will be maintained
- Command line concatenation (see pdfcat in sample code) (by Steve Witham)
- New FAQ; link included in README
- Allow more (although unnecessary) escape sequences
- Prevent exception when reading a null object in decoding parameters
- Corrected error in reading destination types (added a slash since they
are name objects)
- Corrected TypeError in scaleTo() method
- addBookmark() method in PdfFileMerger now returns bookmark (so nested
bookmarks can be created)
- Additions to Sample Code and Sample PDFs
- changes to allow 2up script to work (see sample code) (by Dylan McNamee)
- changes to metadata encoding (by Chris Hiestand)
- New methods for links: addLink() (by Enrico Lambertini) and removeLinks()
- Bugfix to handle nested bookmarks correctly (by Jamie Lentin)
- New methods removeImages() and removeText() available for PdfFileWriter
(by Tien Ha▒)
- Exception handling for illegal characters in Name Objects
Version 1.19, 2013-10-08
------------------------
BUGFIXES:
- Removed pop in sweepIndirectReferences to prevent infinite loop
(provided by ian-su-sirca)
- Fixed bug caused by whitespace when parsing PDFs generated by AutoCad
- Fixed a bug caused by reading a 'null' ASCII value in a dictionary
object (primarily in PDFs generated by AutoCad).
FEATURES:
- Added new folders for PyPDF2 sample code and example PDFs; see README
for each folder
- Added a method for debugging purposes to show current location while
parsing
- Ability to create custom metadata (by jamma313)
- Ability to access and customize document layout and view mode
(by Joshua Arnott)
OTHER:
- Added and corrected some documentation
- Added some more warnings and exception messages
- Removed old test/debugging code
UPCOMING:
- More bugfixes (We have received many problematic PDFs via email, we
will work with them)
- Documentation - It's time for PyPDF2 to get its own documentation
since it has grown much since the original pyPdf
- A FAQ to answer common questions
Upstream changes:
1.000024 2014-06-03 01:52:46Z
- fix inaccurate prerequisite declaration on Test::More (in 1.000023)
1.000023 2014-06-02 02:37:17Z
- support installations on older perls with an ExtUtils::MakeMaker earlier
than 6.63_03
1.000022 - 2014-04-29
- work around change in comparison behaviour in Test::More 0.95_01 by being
more explicit with our tests - now explicitly checking the string form of
the extracted version, rather than the entire version object
- ensure the extracted version is returned as a version object in all cases
(RT#87782, Randy Stauner)
1.000021 - 2014-04-29
- fix use of newer interface from File::Path, to avoid another prereq on
older perls (Graham Knop)
- fixed all out of date prereq declarations
1.000020 - 2014-04-27
- new is_indexable() object method (ether, RT#84357)
- eliminated dependency on IO::File (and by virtue, XS) - thanks, leont!
- removed cruft in test infrastructure left behind from separation from
Module::Build (ether)
- repository moved to https://github.com/Perl-Toolchain-Gang/Module-Metadata
- .pm file is now wholly ascii, for nicer fatpacking (RT#95086)
- some code micro-optimizations (Olivier Mengu)
Upstream changes:
version 1.05: Tue Jun 24 09:38:15 CEST 2014
Fixes:
- test in t/10interp.t failed for Perl 5.20, caused by a
bugfix or change in overload::Overloaded [cpantesters]
version 1.04: Tue Jun 3 10:42:11 CEST 2014
Fixes:
- float serialization under locale in test [cpantesters]
version 1.03: Thu May 22 11:54:24 CEST 2014
Fixes:
- float serialization under locale in test [cpantesters]
- non-errors and ::Dispatcher::Perl
Improvements:
- shorted display of string parameters in stack-trace to max 80 chars
- Log4perl log-lines sometimes show dispatcher as source, skip them.
- disable 'mode switch' trace for try()
version 1.02: Mon Mar 10 16:03:13 CET 2014
Fixes:
- add overload fallback to ::Exception and ::Dispatcher
rt.cpan.org#92970 [Lukas Mai]
- ::Domain::new(translator) with HASH did not initialize
- warn better with ::Optional modules are used before Log::Report
is used.
Improvements:
- changed documentation style
- ::Lexicon::Index dir-scan immediately, hopefully before fork()
Add missing DEPENDS
Upstream changes:
0.04 2014-09-07
- switch to Data::Munge::eval_string
- make with_return a named sub to improve stack traces
- cache runtime generated trampolines; should cut down on the need to
eval strings at runtime
- document RETURN_MULTI_LEVEL_PP
- add RETURN_MULTI_LEVEL_DEBUG for capturing with_return stacks
- Version 3.3.5 fixes a number of bugs in 3.3.4 and earlier versions.
- Version 3.3.4 fixes a small number of bugs in 3.3.3 and earlier versions.
In particular it fixes a bug introduced in 3.3.3 which could cause startup
to fail under some circumstances.
- Version 3.3.3 fixes a small number of bugs in 3.3.2 and earlier versions.
In particular it fixes a bug introduced in 3.3.2 which could cause
logging to stop under some circumstances.
Upstream changes:
MediaWiki 1.23.5
This is a security release of the MediaWiki 1.23 branch.
Changes since 1.23.4
(bug 70672) SECURITY: OutputPage: Remove separation of css and js module allowance.
Changes in 0.5.2
fcdf0f8 Autoconf: check for htonll as declaration in a header file
5790ec7 SSL: correctly report hostname verification errors.
d60c28c Build: disable OpenSSL deprecation warnings on OSX
072191a Lib: include platform, version and copyright in AMQP handshake
8b448c6 Examples: print message body in amqp[s]_listen[q] examples
7188e5d Tools: Add flag to set prefetch for amqp-consume tool
Changes in 0.5.1
a566929 SSL: Add support for wildcards in hostname verification (Mike Steinert)
a78aa8a Lib: Use poll(2) instead of select(2) for timeouts on sockets.
357bdb3 Lib: support for specifying frame and decoding pool sizes. (Mike Stitt)
8956003 Lib: improve invalid frame detection code.
b852f84 Lib: Add missing amqp_get_server_properties() function.
7001e82 Lib: Add missing ssize_t on Win32 (emazv72)
c2ce2cb Lib: Correctly specify WINVER on Win32 when unspecified.
fe844e4 CMake: specify -DHAVE_CONFIG_H in examples.
932de5f Lib: correct time computation on Win32 (jestor)
3e83192 HPUX: use gethrtime on HP-UX for timers.
cb1b44e HPUX: correct include location of sys/uio.h
8ce585d Lib: incorrect OOM condition when 0-lenth exchange name is received.
c7716b8 CMake: correct htonll detection code on platforms defined with a macro.
4dc4eda Lib: remove unused assignment.
45302cf Lib: remove range-check of channel-ids.
A new libmongoc-priv.so library is installed that does not have symbols
hidden. You can access private headers via the -private.h variants. This
means you will need to recompile your project every time the library is
changed (if you use those private headers, as they are subject to change).
For other minor changes, see commits made since 0.98.0:
https://github.com/mongodb/mongo-c-driver/compare/0.98.0...1.0.0
LGOGDownloader 2.18
- Added better login check
* Allows user to login with username instead of email
* Prints some info on failed login with debug build
- Fixed login issues caused by case sensitivity
- Added PREFIX and MANPREFIX to Makefile
- Allowed unrecognized options in config file
- Added --no-cover option back
- Listing games with --list now only displays games that support the platforms specified with --platform
- Made Downloader::HTTP_Login print some more info
- File hashing changes to Util::createXML
* File hash is updated in the same loop that chunk hashes are calculated (gives the user better indication of the hashing progress)
- Avoid touching $HOME when displaying help or version (patch by: Stephen Kitt)
* This fixes main() to handle --help and --version before doing anything to the filesystem
- Fixed spelling mistake in message displayed when downloading files (patch by: Neil Ramsbottom)
- Changed --report option to allow setting the filename of report
- Old files are now renamed instead of deleted during --repair --download
- Renamed old files now contain date in filename (filename.ext.YYYYMMDDTHHMMSS.old where T is the date-time separator)
- Added support for setting subdirectories
* Allows user to specify subdirectories for games, installers, extras, patches, language packs and dlc
* You can use templates "%gamename%", "%platform%" and "%dlcname%" which are replaced by the relevant info
* New options: --subdir-installers, --subdir-extras, --subdir-patches, --subdir-language-packs, --subdir-dlc, --subdir-game
r123:
Added : experimental lz4frame API, thanks to Takayuki Matsuoka and
Christopher Jackson for testings
Fix : s390x support, thanks to Nobuhiro Iwamatsu
Fix : test mode (-t) no longer requires confirmation, thanks to Thary Nguyen
r122:
Fix : AIX & AIX64 support (SamG)
Fix : mips 64-bits support (lew van)
Added : Examples directory, using code examples from Takayuki Matsuoka
Updated : Framing specification, to v1.4.1
Updated : xxHash, to r36
r121:
Added : Makefile : install for kFreeBSD and Hurd (Nobuhiro Iwamatsu)
Fix : Makefile : install for OS-X and BSD, thanks to Takayuki Matsuoka
