This includes a lot of changes - see the internal CHANGES.md file for
details. Also included is support for OCaml 4.14, as well as a revamp
of the internal structure and dependencies.
Bug Fixes
-dependency update; upgrade to trash v2.1.1 .
The trash upgrade makes sure that trashed items on mount points on freedesktop
are actually restorable.
There are many changes since the last version, including a
from-the-ground-up rewrite of the parser generator code and a switch of
the build system to dune. More details in the CHANGES.md file included
in the distribution.
3.3.0
* Features
* Add CLI flag `--suffix_list_url` to set the suffix list URL(s) or source file(s)
* Add CLI flag `--no_fallback_to_snapshot` to not fall back to the snapshot
* Add alias `--include_psl_private_domains` for CLI flag `--private_domains`
* Bugfixes
* Handle more internationalized domain name dots
* Misc.
* Update bundled snapshot
* Add basic CLI test coverage
0.99 (2022-01-08)
Added: Python 3.9 & 3.10 are now supported
Added: Sphinx 4 is now supported.
Added: Can now set background images using PageBreak.
Added: Can now set multiple style names in the class directive
Added: pdf_smartquotes option is now supported with Sphinx builds
Added: Add support for styling separations.
Changed: Outline entries that are deeper than the ToC are now collapsed
Changed: math directive updated to support latest matplotlib
Changed: the --very-verbose option provides more information when things go wrong.
Changed: The default stylesheet has been improved. Use the rst2pdf-0.9 stylesheet to apply the previous default as per the "Migrating to the New Default Stylesheet" section of the manual. Also, the twoColumn style is no longer part of the default styles and is now a separate style. Add -s twocolumn to apply it or, for Sphinx add pdf_stylesheets = ['twocolumn']
Changed: Stylesheets are now in YAML. Migrate to the new format using python -m rst2pdf.style2yaml as per the "Migrating Stylesheet Format" section of the manual.
Fixed: An imagine that is too tall in a table cell is now resized to fit.
Fixed: rst2pdf now writes to stdout again
Fixed: Ensure table columns are wide enough for use
Fixed: Better error messages for malformed RST
Fixed: The Sphinx versionadded & versionchanged directives work again
Fixed: Hyperlinks with embedded internal link aliases no longer crash.
Fixed: A more sensible error message is displayed when importing an extension fails
Fixed: If rst2pdf errors, it should now return a non-zero status code.
Fixed: Various documentation improvements
Fixed: Addressed regular expression FutureWarning.
Removed: Documentation related to kerning has been removed as it depended on word-axe which is longer
This is a bugfix release. The most notable changes are: more work on
fixup command that was introduced in the previous release, and a couple
of bug fixes. Evolve is no longer confused when trying to select the
latest successor in cases where a changeset was split and its child was
pruned, see https://bz.mercurial-scm.org/show_bug.cgi?id=6648. Evolve
with --dry-run flag used to update working directory in certain cases,
see https://bz.mercurial-scm.org/show_bug.cgi?id=6669, this was fixed.
There are also changes that should help Debian package maintainers; e.g.
it's now possible to use system cbor instead of the one we vendor.
= Mercurial 6.1.2 =
* Improve Windows test suite
* Fix `debuglock` not ignoring a missing lockfile when forcing a lock
* Improve help of `ui.large-file-limit`
* Set the large-file-limit to 10MB (from 10MiB) for clarity
* While rewriting desc hashes, ignore ambiguous prefix "hashes"
* Fix a crash in partial amend with copies
* Fix a py3 compatiblity bug
* Fix incorrect metadata causing dirstate-v2 data loss in edge case
* Fix cleanup of old dirstate-v2 data files when using `rhg`
* Make reference to `.hg/last_message.txt` relative in commit
* Fix an infinite hang when `rhg` is used in the background
* Fix Python DLL loading bug in Windows
* Add `--docket` flag to `debugstate` to check out dirstate-v2 metadata
* Remove `debugdirstateignorepatternhash` in favor of `debugstate --docket`
* Fix incorrect metadata causing systematic complete dirstate-v2 rewrite
pkgsrc changes:
---------------
* Create ${VARBASE}/run directory since smartd does not start if it does not
exist.
* Add ${VARBASE}/run existence test in the rc.d script.
upstream changes:
-----------------
* Date 2022-02-28
Summary: smartmontools release 7.3
[https://www.smartmontools.org/browser/tags/RELEASE_7_3/smartmontools/NEWS]
Merged
autotools: Add ws2_32 library with MagickCore for Win32 platform #5119
Avoid NULL pointer dereference in coders/wmf.c #5117
Commits
beta release 600c455
possible null dereference 4e085a2
improved error checking 2e5c976
Also create an arm64 installer. c84fcf5
Also create portable arm64 binaries. 6126915
eliminate coverity defect 4cec11d
eliminate Coverity defect 0364622
eliminate Coverity defect a619616
ImageMagick/ImageMagick#5099 43da986
prevent memory leak on exception 0157eb8
the tiles per page cannot be zero fecf247
check security policy before reading an -fx expression from a file c53fb96
fix exception on a valid ternary @ ImageMagick/ImageMagick#4533 64b5fe6
display the image pixel cache type 759029d
change order of pixel cache type 3529df8
change order of pixel cache type bacb357
use the correct property key for arithmetic coding 0184b6c
Reverted patch because HDRI should always be highres. 4e20a27
the maximum dissolve factor is 1.0 452c436
Removed duplicate check. b108dbd
Fixed possible memory leak reported in #5121. 9c957a6
Removed to heif_filetype_yes_unsupported check to resolved the issue reported in #5123. 47898c6
possible memory leak @ ImageMagick/ImageMagick#5121 b0c1967
eliminate compiler warnings a10a570
eliminate compiler warnings 3fa3d9f
conditional logging 53fb342
eliminate compiler warnings a429f44
background kept when making cylinder @ ImageMagick/ImageMagick#5112 925a6c5
cosmetic d4d5ca2
possible memory leak @ ImageMagick/ImageMagick#5131 8ed6d1d
release 4e6f7b1
release 39baedd
The major new features:
Multi-barcode reading with the new std::vector<Result>
ReadBardcodes(...) function. Does not work for Aztec, DataMatrix
and Maxicode, yet.
Multi-resolution scanning that can automatically downscale the
input to substantially increase the detection rate on high
resolution scans, see DecodeHints::tryDownscale, currently only
enabled in the ReadBardcodes function
New Result::symbologyIdentifier property (thanks to @gitlost)
Updated and improved android wrapper (thanks to @markusfisch)
Worth mentioning:
Most deprecated APIs from last release were removed (will likely
break your build if you still have deprecation warnings with
1.2.0)
New Result::lineCount and DecodeHints::minLineCount properties
to improve reliability of 1D code reading
New Result::isMirrored property (works for QRCode, PDF417,
DataMatrix)
Improved FNC1 handling and deprecation of DecodeHints::assumeGS1
Lots of small fixes all over the place
Drop in-tree copy of stb image library
New OpenCV example code
QRCode reading improvements
Python 3.10 support
Major changes between OpenSSL 1.1.1n and OpenSSL 1.1.1o [3 May 2022]
o Fixed a bug in the c_rehash script which was not properly sanitising
shell metacharacters to prevent command injection (CVE-2022-1292)
2.4.16
Changed
- Make `fs.zipfs._ZipExtFile` use the seeking mechanism implemented
in the Python standard library in Python version 3.7 and later
- Mark `fs.zipfs.ReadZipFS` as a case-sensitive filesystem
- Optimized moving files between filesystems with syspaths.
- Fixed `fs.move.move_file` to clean up the copy on the destination in case of errors.
- `fs.opener.manage_fs` with `writeable=True` will now raise a `ResourceReadOnly`
exception if the managed filesystem is not writeable.
- Marked filesystems wrapped with `fs.wrap.WrapReadOnly` as read-only.
v2.11.5
Fix crash while obtaining object_type() of an Unknown node.
Fix a bug where in attempting to handle the patching of distutils by virtualenv,
library submodules called distutils (e.g. numpy.distutils) were included also.
hck is a shortening of hack, a rougher form of cut.
A close to drop in replacement for cut that can use a regex delimiter instead
of a fixed string. Additionally this tool allows for specification of the order
of the output columns using the same column selection syntax as cut.
No single feature of hck on its own makes it stand out over awk, cut, xsv or
other such tools. Where hck excels is making common things easy, such as
reordering output fields, or splitting records on a weird delimiter. It is
meant to be simple and easy to use while exploring datasets. Think of this as
filling a gap between cut and awk.
