ChangeLog:
Features:
- The unsparsify -f feature
- The new sort-within-records verb is an old ask, underway from the Go
port, backported to C
- Likewise the truncate DSL function
Bugfixes:
- The count -n feature was not implemented as intended
- Pretty-print format now works correctly with --headerless-csv-output
- The seqgen verb now correctly tracks NR and FNR in the records it emits
- An intermittent JSON-parsing bug has been fixed
v0.4.2
Use urlparse from urllib.parse (Python 3) or urlparse (Python 2)
to prepare for the deprecation of urllib.parse.splittype.
Unix socket clients now send localhost as Host: HTTP field instead of
the path to the socket
Added a TransportError exception, subclass of ProtocolError, which
provides more details
Added PowerPC 64 architecture (ppc64le) to Travis CI runs, to ease the
integration of new release into RHEL/Ubuntu
- Add exception-related files which were new in 1.2.25.3 to MANIFEST
so they're shipped. Reported by Tina Müller and Leah Neukirchen.
- Add metadata links to homepage, bugtracker and VCS. Suggested by
Gabor Szabo.
- Allow AUTOMATED_TESTING builds against xapian-core 1.4.x.
- Properly handle newer exceptions. The C++ exception class
WildcardError wasn't known to the Perl bindings, and would result in
croak("something terrible happened"). The C++ exception classes
DatabaseClosedError and DatabaseNotFoundError also weren't known, but
these would instead be reported as their base classes DatabaseError
and DatabaseOpeningError respectively. Reported by Adam Sjøgren.
- README: Update to reflect that the SWIG-generated Perl bindings are
stable since xapian-bindings 1.4.16.
pugixml-1.11 is out. This is a maintenance release that fixes some
compatibility issues and introduces several new APIs (read the
changelog for details).
Highlights for this release include:
* Add xml_node::remove_attributes and xml_node::remove_children
* Add a way to customize floating point precision via xml_attribute::set
and xml_text::set overloads
* XPath parser now limits recursion depth which prevents stack
overflow on malicious queries
Changes since 0.16.0:
v0.17.1
Bugfixes
* Running bat without arguments fails ("output file is also an input"), see #1396
v0.17.0
Features
* Added a new --style value, rule, which adds a simple horizontal ruled line
between files, see #1276 (@tommilligan)
* Pass -S ("chop long lines") to less if --wrap=never is set in bat, see
#1255 (@gahag)
Bugfixes
* Detect infinite loop when input and output are the same, see #1193 and
#1197 (@niklasmohrin)
* Throw an error when bat is being used as pager, see #1343 (@adrian-rivera)
* Bash syntax highlighting not selected for *.ebuild and *.eclass files, see
#1292 (@sharkdp)
* Fix zsh completion when using -p, see #1320 (@xzfc)
Other
* Add note to refer to see detailed help with --help (and vice versa with
-h), see #1215 (@henil)
* Add a Contributors section to README, see #1348 (@adrian-rivera)
Syntaxes
* Manpage syntax highlighting has been improved, see #1315 (@keith-hall)
* Add Svelte file syntax, see #1285 (@kjmph)
New themes
* Coldark, see #1329 (@ArmandPhilippot)
CLI tool to perform spell checking.
Rust implementation of Peter Norvig's Spell Corrector.
The default file contains ~30k unique words and is included in the crate.
Currently stava only supports the English alphabet.
A bunch of packages had an extra "p" specification character passed to
the pax -p option. One is enough. Committed to reduce the human parsing
costs, should someone else need to examine this. (In my case because it
seems recent Linux distros have changed such that some -p arguments can
now cause an error to occur, where previously they were accepted.)
4.6.2
Bugs fixed
* A vulnerability (CVE-2020-27783) was discovered in the HTML Cleaner by Yaniv Nizry,
which allowed JavaScript to pass through. The cleaner now removes more sneaky
"style" content.
Reviewing changes involves reading diffs. Sometimes, a line-oriented
presentation of changes is not precise enough, especially when changes
involve long lines or very similar consecutive lines.
This program processes such diffs, and output them (in the console)
with additional diff information on top of the unified diff format,
using text attributes.
It works hunk by hunk, recomputing the diff on a word-by-word basis.
1.4.3
- No longer use a universal wheel
1.4.2
- Fix exception raised when calling `PyQuery("<textarea></textarea>").text()`
- python2 is no longer supported
7.1.3
Fixed handling of buffer boundaries in format_to_n.
Fixed linkage errors when linking with a shared library.
Reintroduced ostream support to range formatters.
Worked around an issue with mixing std versions in gcc
7.1.2
Fixed floating point formatting with large precision.
7.1.1
Fixed ABI compatibility with 7.0.x.
Added the FMT_ARM_ABI_COMPATIBILITY macro to work around ABI incompatibility between GCC and Clang on ARM.
Worked around a SFINAE bug in GCC 8.
Fixed linkage errors when building with GCC's LTO.
Fixed a compilation error when building without __builtin_clz or equivalent.
Fixed a sign conversion warning.
7.1.0
Switched from Grisu3 to Dragonbox for the default floating-point formatting which gives the shortest decimal representation with round-trip guarantee and correct rounding.
Added an experimental unsynchronized file output API which, together with format string compilation, can give 5-9 times speed up compared to fprintf on common platforms.
Added a formatter for std::chrono::time_point<system_clock>.
Added support for ranges with non-const begin/end to fmt::join.
Added a memory_buffer::append overload that takes a range.
Improved handling of single code units in FMT_COMPILE.
Added dynamic width support to format string compilation.
Improved error reporting for unformattable types: now you'll get the type name directly in the error message instead of the note.
Added the make_args_checked function template that allows you to write formatting functions with compile-time format string checks and avoid binary code bloat.
Replaced snprintf fallback with a faster internal IEEE 754 float and double formatter for arbitrary precision.
Made format_to_n and formatted_size part of the core API.
Added fmt::format_to_n overload with format string compilation.
Added fmt::format_to overload that take text_style.
Made the # specifier emit trailing zeros in addition to the decimal point.
Changed the default floating point format to not include .0 for consistency with std::format and std::to_chars. It is possible to get the decimal point and trailing zero with the # specifier.
Fixed an issue with floating-point formatting that could result in addition of a non-significant trailing zero in rare cases e.g. 1.00e-34 instead of 1.0e-34.
Made fmt::to_string fallback on ostream insertion operator if the formatter specialization is not provided.
Added support for the append mode to the experimental file API and improved fcntl.h detection.
Fixed handling of types that have both an implicit conversion operator and an overloaded ostream insertion operator.
Fixed a slicing issue in an internal iterator type.
Fixed an issue in locale-specific integer formatting.
Fixed handling of exotic code unit types.
Improved FMT_ALWAYS_INLINE.
Removed dependency on windows.h.
Optimized counting of decimal digits on MSVC.
Improved documentation.
Added the FMT_REDUCE_INT_INSTANTIATIONS CMake option that reduces the binary code size at the cost of some integer formatting performance. This can be useful for extremely memory-constrained embedded systems.
Added the FMT_USE_INLINE_NAMESPACES macro to control usage of inline namespaces.
Improved build configuration
2019/11/29 (diffstat 1.63)
+ eliminate fixed buffer when decoding range.
+ use locale in computing filename column-width.
+ improve parsing for git diffs.
+ use terminal-width as default for -w to tty.
+ minor fix in do_merging (Miloslaw Smyk).
+ improve relative-pathname matching in count_lines()
+ add a parsing-case for svn diff.
+ quote filenames in -t/-T output.
+ fix cppcheck warnings about sscanf.
+ update configure macros
+ update config.guess, config.sub
serd (0.30.6) stable;
* Fix various minor warnings and other code quality issues
* Improve reader error handling
* Make all free functions tolerant of NULL
* Parse prefixed names more strictly
-- David Robillard <d@drobilla.net> Sun, 27 Sep 2020 10:51:38 +0000
sord (0.16.6) stable;
* Fix potential crash or incorrectness issues with GCC 10
* Fix various minor warnings and other code quality issues
-- David Robillard <d@drobilla.net> Sun, 27 Sep 2020 11:08:18 +0000
New v3: performance enhancements and other improvements with a redesigned search engine. Search speeds are significantly increased when searching very large files (several GB). Added a new T3 benchmark with a large 13GB file to demonstrate the improvements. More features will be added soon!
zoxide is a blazing fast alternative to cd, inspired by z and z.lua.
It keeps track of the directories you use most frequently, and uses a ranking
algorithm to navigate to the best match.
Upstream changes:
=== Version 1.001 (2020-11-05)
lib/PDF/Table.pm fix edge case where no borders led to no rules [GH 55].
t/manifest.t, t/pod.t
Make author-only (AUTHOR_TESTING=1). ref [GH 61]
lib/PDF/Table.pm, t/pod.t
POD change =head5 to =head4, to look better and stop older POD checkers
from complaining (in t/pod.t). Also minimum Test::Pod version 1.52
(was 1.00). ref [GH 61].
examples/sample1.pl, t/lib/TestData.pm, INFO/Deprecated, lib/PDF/Table.pm
Replace deprecated setting names by current names in the POD, examples, and
t-tests. After November 2022, settings names (args) with a leading dash
(hyphen) will no longer be permitted.
=== Version 1.000 (2020-11-03)
INFO/Deprecated, INFO/Table.html, lib/PDF/Table.pm
Final cleanup of POD and generation of HTML documentation, and listing of
all deprecated setting names.
t/lib/PDFAPI2Mock.pm add linedash dummy call so t tests can run.
lib/PDF/Table.pm, MANIFEST, examples/border_rules.pl, util/3_examples.pl
Old "borders" was vertical and horizontal rules PLUS outside rules (frame).
Separate into "borders" (outside) and "rules" (inside). Add border_rules.pl
example to illustrate.
examples/chess.pl with DejaVu-Sans, can use chess piece images
.github/worksflows/test.yml, README.md minor fixes in README to make lint
run happily, stop checking Table.html.
examples/*.pl Allow -A or -B on the command line (case insensitive) to
override PDFpref file (if any) setting. This speeds up testing for PDF::API2
vs PDF::Builder (default).
lib/PDF/Table.pm, INFO/Deprecated, INFO/Table.html, MANIFEST,
examples/chess.pl, t/PDF-Table.t, t/lib/PDFAPI2Mock.pm, util/3_examples.pl
Clean up handling of minimum and maximum cell widths (min_w and max_w).
Clean up handling of background and foreground colors. Add chess.pl example
to show off color handling. Update t-tests for slight coordinate changes.
Add row_props (row properties) similar to column_props, to set things on a
per-row basis.
If the table spills over to a new page, and next_y and/or next_h was NOT
given, issue a warning and for a default use 90% of the media height for
next_y and 80% for next_h (ref ticket #11). This should be better than
using the (start_)y and (start_)h values for the table.
Add new, more consistent aliases for a number of settings (ref ticket #13).
The old setting names are still available for at least TWO YEARS, after
which, they will be REMOVED. Over the next two years you should update your
code to use the new setting names. The settings involved are:
'start_y' is now 'y'
'start_h' is now 'h'
'row_height' is now 'min_rh'
'background_color' is now 'bg_color'
'background_color_odd' is now 'bg_color_odd'
'background_color_even' is now 'bg_color_even'
'font_color' is now 'fg_color'
'font_color_odd' is now 'fg_color_odd'
'font_color_even' is now 'fg_color_even'
'font_underline' is now 'underline'
'lead' is now 'leading'
These are listed in INFO/Deprecated. In the near future, the examples and
t-tests will be updated to use the new setting names.
Ship with Table.html (documentation) in INFO/. Update MANIFEST.
Remove 010_fonts.pl from the util/3_examples.pl list. I'm still not sure why
it should be in the examples -- basically, it shows that UTF-8 doesn't work
with core fonts, but that's already a documented limitation (in
PDF::Builder). 010_fonts.pl is still in the examples/ directory, if anyone
wants to play with it.
Add INFO/Deprecated to list names and interfaces scheduled to be removed in
the future (at least 2 years). MANIFEST updated.
Split out older Changes list into INFO/Changes_2019. MANIFEST updated.
Increase default cell padding to 2pt. See documentation "COMPATIBILITY" for
how to get the old behavior (0pt padding). t/Basics.t, t/Colspan.t,
t/PDF-Table.t expected write locations updated. examples/colspan.pl and
examples/header.pl were widened slightly to accommodate the padding.
Make "odd/even" row consistent with or without a header (repeated or not),
particularly when a table is split across pages, and when a row itself is
split in such a case. See documentation "COMPATIBILITY" for how to get the
old behavior.
Header 'repeat' default changed from 0 (no) to 1 (yes), as most of the time
a user will want to repeat any header on each new page. Suppress repeating
a header with a value of 0. See documentation "COMPATIBILITY" for how to
get the old behavior.
Currently, PDF::Table is shipped to use the new behaviors, with as much
compatibility as possible with existing code (see $compat_mode in Table.pm).
Unless you have need to use the old behaviors for existing code, we
suggest that you leave $compat_mode as 0, use the 'compatibility' setting, or
individually set the flags, to make use of the new behaviors. Also use the
new settings names, unless there is a need for compatibility in existing
code. Eventually, use of the old names will produce warnings, and later,
error messages, but the old behaviors should be available for a long time.
You should examine your existing code and plan to update it over time.
Initialize %arg entries (defaults) so no uninitialized errors. #57 reported by
Larry Leszczynski (larryl) for text_block() along with PR.
Minor build process cleanup
Minor spelling and typo fixes in POD and examples
v1.3.0:
* Drop support for Python 3.5
* Add XmlDocument and Wsdl11Document classes
* Refactoring of XMLResource to support ElementTree-like XPath API
on both full and lazy modes
ugrep v3.0.4
New v3: performance enhancements and other improvements with a
redesigned search engine. Search speeds are significantly increased
when searching very large files (several GB). Added a new T3
benchmark with a large 13GB file to demonstrate the improvements.
Update v3.0.2 corrects an issue with ugrep v3.0.0 when opening
root-owned files on some Linux systems and resolves a minor problem
when matching a specific short pattern. Update v3.0.4 includes
improvements and a fix for an issue with large zip files searched
in the query UI and corrects a compressed file search issue on
Windows systems. More features will be added soon!
ugrep v2.5.6
Interactive query UI usability improvements; new query UI commands
Alt-[ (decrease fuzziness by one), Alt-] (increase fuzziness by
one), Alt-{ (decrease context size by one line), Alt-} (increase
context size by one line); improved ARM AArch64 configure and build.
More coming soon!
2.2.13 (November 2, 2020)
-------------------------
Make the enchant program output its version to standard output, not standard
error. This may help some programs that use this output.
Version 9.0.4 (2020-10-20)
--------------------------
.Bug fixes
- Fix listing out installed plugins (e.g. --filter list)
- Fix python version check failing on 3.10 (thanks @hoadlck)
.Testing
- Update to deadsnakes/python@v2.0.0 for testing dev python versions
- Move from testing against 3.9-dev to stable 3.9
- Add 3.10-dev test target
1.62 2020-10-15
- dist.ini / weaver.ini / .tidyallrc / etc. cleanup
- Move to @SHLOMIF
- Inspired by Lady_Aleena.
- Sympotamtic fix for item/link nested inside custom tags:
- https://github.com/shlomif/perl-XML-RSS/issues/7
- Thanks to @jkramer for the bug report and a test case.
