Upstream changes:
0.39 - 2013-08-17, H.Merijn Brand <h.m.brand@xs4all.nl>
* Upped copyright to 2013
* Add alternative test results for 5.19.3 and up (COW and no PV for IV)
Upstream changes:
2.35
Apply patches by Ilmari and D. Steinbrunner.
Should work on 5.18 now.
2.34
Administrative release to eliminate build dependency on module
used only for testing.
2.33
Caught up with various RT reports, and fixed a bunch of them.
[RT #74198] fixed dumps containing read-only objects (obtained via JSON::XS)
[RT #74198], Fixed stringification of code compiled under strict in later
perls. Thanks to Father Chrysotomos for the patch and nudge.
Also eliminated a false-fail prone test which was actually pretty useless,
which should close several tickets which I wont bother to list.
This release also is the first from the new git repository at
https://github.com/demerphq/Data-Dump-Streamer
Upstream changes:
2013-05-10 Gisle Aas <gisle@ActiveState.com>
Release 1.22
Deal with class names containing '=' [RT#72415]
Filters now get access to $ctx->expr; the textual expression
denoting the current object.
Upstream changes:
1.23 2013-09-26 DCANTRELL Add David Muir Sharnoff's tests for
really big data structures (but not yet working);
Check for taint-mode less insanely, thanks to Ilmari;
Don't delay loading File::Find::Rule (see RT 87554)
Upstream changes:
version 1.18; 2013-09-21
* bugfix: store peep chain link reliably under threads
* update tests for Perl 5.19.4's removal of the special treatment of
aliased undef in arrays
* in doc, switch to consistent use of British English spellings
* doc typo fix
version 1.17; 2013-08-04
* bugfix: correct the dtrace-related macro squashing for Perls prior
to 5.13.8, where the macros had the wrong number of parameters and
prevented compilation
* bugfix: use core's PL_no_localize_ref string constant as data string
rather than format string
* update to handle the new padrange op type in Perl 5.17.6
* update to handle the new op allocation mechanism in Perl 5.17.2
* clarify and expand documentation around list-like assignments
* clarify documentation of deref() regarding lvalueness
* remove the bulk of the documentation from the README file
Upstream changes:
0.004004 Sun Aug 5 17:46:56 2012
- Further patch to failures.t
- Added redefinition of Scalar::Util::blessed() to avoid nasty
surprises (thanks Andrew!)
- Added confess() and cluck() overrides
and fixed caller() override to set @DB::args when appropriate
0.004005 Mon Aug 20 10:44:35 2012
- Further patch for t/confess.t
(Thanks, David!)
0.004006 Thu Oct 4 16:49:47 2012
- Circumvented annoying undef-as-key warnings in Perl 5.17+
- Honour @CARP_NOT's when reporting context
0.004007 Fri Oct 5 23:05:05 2012
- Added BLESSED handler for better control over how blessed() lies
- Upgraded reimplementation of blessed() to make more sense
Upstream changes:
2.20 Sun Nov 24 10:52:00 2013
- Relax pre-req version requirements
2.19 Sun Sep 15 09:16:00 2013
- Change VERSION => $VERSION in Makefile.PL to VERSION_FROM => 'lib/Config/Tiny.pm'.
Reported by Jean-Louis Morel. See RT#88670.
2.18 Sat Sep 14 10:03:00 2013
- Remove obsolete and wrong version # from Makefile.PL.
Reported by Jean-Louis Morel. See RT#88658.
- Implement Kevin Ryde's suggestion to test if read() will return undef.
If so, set an error message and (still) return undef.
2.17 Fri Sep 13 12:41:00 2013
- Remove the file tests -efr during calls to read(). The open() tests for any error.
Also, the -f test was reporting /dev/null as a directory, not a file.
Thanx to Kevin Ryde for pushing me to implement this. See RT#36974.
- Clean up some error messages slightly.
2.16 Fri Sep 6 11:54:00 2013
- Replace Path::Tiny with File::Spec, because the former's list of dependencies is soooo long :-(.
Changed files: t/02.main.t, t/04.utf8.t, Build.PL and Makefile.PL.
See: RT#88435 (for Tree::DAG_Node) for an explanation.
2.15 Sun Aug 4 14:59:00 2013
- Clean up the shambolic dates in this file.
- Add a note under Caveats about setting options more that once. Only the first case is respected.
Thanx to Kimmel K. See RT#69795.
- Add a $encoding parameter to read_file() and write_file(). See docs for details.
Add t/04.utf8.t and t/04.utf8.txt.
Thanx to Mark Lawrence and Wolfgang Husmann. See RT#71029 and RT#85571.
- For BSD-based systems, when writing a file during tests, use:
my($temp_dir) = File::Temp -> newdir('temp.XXXX', CLEANUP => 1, EXLOCK => 0, TMPDIR => 1);
- Rename t/*.t files. I use '.' rather than '_' in file names because the latter is a shift char.
- Add MANIFEST.SKIP, Changelog.ini, Build.PL, META.json.
- Add an FAQ to the docs.
- Clean up the docs.
Upstream changes:
1.68 Mon Aug 6 22:22:22 PDT 2012
Patch fromFileGetTopLines() and fromFileIsHeader(), which are used by fromFile(). Impact: minor.
Improve performance of fromFileGuessOS()
Improve fromFile(), fromCSV(), csv() to support using \r, \n within a CSV field.
join() now supports {matchNULL => 1, NULLasEmpty => 1}, if one would like to treat NULL as empty string, or
treat NULL as equal (however, not equal to empty string). Both are set to 0 by default.
Suggested by Kyle Horton & Wilson Dave.
Remove inheritance from AutoLoader and Exporter.
Thanks to Brian Wightman
Thanks to Nicholas Andonakis for sharing his code, quite a few ideas in his package inspired the improvements below!
Add new shortcut methods: lastRow(), lastCol(), colName($colNumericIndex)
One can now write
foreach my $i (0..$t->lastRow)
instead of
foreach my $i (0..$t->nofRow-1)
Add iterator(), so that one can now write
my $next = $t_product->iterator();
while (my $row = $next->()) {
# have access to a row as a hash reference, access row number by &$next(1);
$t_product->setElm($next->(1), 'ProductName', 'New! '.$row->{ProductName});
}
addCol() can take the default value for the new column (first argument)
addRow() supports {addNewCol => 1}
moveCol() can take a $newColName.
setElm() can set a value for multiple cells, specified by ref to row array and col array
match_string(), match_pattern(), match_pattern_hash() also produce $parentTable->{MATCH}
# match returns all matched row ids in $t_product->{MATCH} (ref to row ID array)
$t_product->match_pattern_hash('$_{UnitPrice} > 20');
# create a new column, with 'No' as the default value
$t_product->addCol('No', 'IsExpensive');
# use $t_product->{MATCH} to set values for multiple Elements
$t_product->setElm($t_product->{MATCH}, 'IsExpensive', 'Yes');
1.67 Wed Jul 25 11:47:23 PDT 2012
Update Change.txt file to point out $keepRestCol defaults to 1 is only for group()
For pivot(), $keepRestCol is still default to 0 as before.
1.66 Wed Jul 25 11:03:29 PDT 2012
Change the default value of keepRestCol in group() to 1, instead of 0 to be compatible with older versions
Thanks to Kyle Horton
Upstream changes:
1.08 2013.12.11
Eliminate pipe from CGI test script t/09
Filehandle => FileHandle typo fixed#91312
1.07 2013.12.10
New maintainer SZABGAB
Tried to reconstruct entries in the CHANGES file
Some more test case
Code clean-up (remove indirect calls)
Remove Build.PL
use warnings;
Officially require perl 5.006.
Move example file to the eg/ directory
mga release to fix XAA loading and running.
Alan Coopersmith (1):
Remove mga_esc.c vestigal stub
Andy MacLean (1):
Unbreak dual head (xinerama)
Dave Airlie (2):
mga: fix XAA runtime
mga: bump release to 1.6.3
Egbert Eich (3):
Disable Linear Expansion on BE
Restructure code to make it more readable
Disable HW cursor by default on G200 server chips
Julien Cristau (1):
Restrict kernel driver matching to g200 server chips
Tormod Volden (2):
Fix dual head crash without XAA
mga: Use exaGetPixmapFirstPixel() instead of devPrivate.ptr
