[libdbd-pg-perl] annotated tag 3.0.0 created (now d33611e)
Christoph Berg
myon at debian.org
Thu May 25 10:55:15 UTC 2017
This is an automated email from the git hooks/post-receive script.
myon pushed a change to annotated tag 3.0.0
in repository libdbd-pg-perl.
at d33611e (tag)
tagging 266811523151e25edb09cfd6a612a3f31217a46a (commit)
replaces 2.19.3
tagged by Greg Sabino Mullane
on Tue Feb 4 16:38:44 2014 -0500
- Log -----------------------------------------------------------------
Version 3.0.0, released February 3, 2014
-----BEGIN PGP SIGNATURE-----
iEYEABEDAAYFAlLxXeQACgkQvJuQZxSWSshNhwCg1BL/eoAnvwkUWTYsBgs948r1
7VgAn3Z3nfyrtJWdf40gKfKh/6zinXs5
=wqPZ
-----END PGP SIGNATURE-----
Bryan Carpenter (1):
Fix typo in docs regarding pg_placeholder_dollaronly. RT #91400
Dagfinn Ilmari Mannsåker (6):
Fix ORDINAL_POSITION in foreign_key_info. Per RT 88794
Fix foreign_key_info with unspecified schema
Use correct SQL_BIGINT constant for int8
Fix assertion when binding array columns on debug perls >= 5.16
Avoid reinstalling driver methods in threads. Per RT #83638
Explicitly specify en_US for spell checking
David Dick (1):
Add support for AutoInactiveDestroy.
David E. Wheeler (2):
Print instead of prompt when running unattended.
Update App::Info.
Erik Rijkers (1):
Use DEBUG1 for the PrintWarn test, as Postgres got less chatty in 9.3
Fitz Elliott (1):
cleanup dangling largeobjects in tests
Greg Sabino Mullane (117):
Signature for 2.19.3
Spelling correction per #78168
Note git commits in recent version releases.
Attempt to output some compiler information in case DBD::Pg fails to load at all,
Mention the spclocation/pg_tablespace_location() change for 2.19.3
Note bug 78168
Make sure CREATE TABLE .. AS SELECT returns rows in non do() cases.
Credit last change
Bump to 2013
Tweak the test for lo_import_with_oid requirement yet again. RT bug 83145
Better AutoCommit docs. RT 82356
Fix previous PGLIBVERSION check.
Remove math.h requirement: should not be needed anymore, and causing problems on some platforms. RT #79256
Note bug number
Mention hstore to hash mapping
Typo fixes per rt.cpan.org #86378
Give info on mailing list subscription and archives.
Soecify exactly what .pl and .pm files we want copied to the lib dirs,
Do not assume oid will fit in an int! Per RT 85836
Note last change
Add new methods for pg_resetxlog and pg_upgrade. Bump copyright.
For testing setup:
Another stab at the UTF-8 system, this time simplified as much as possible.
Typo
Make pg_enable_utf8 accept ints, and reject invalid values.
Put a skeleton utf8 test place.
Update types via 9.2 source tree
Confirmed that lo_import_with_oid does not exists anywhere in the 8.4 source. Why is this giving us so many problems?
Make sure we respect our "canonical SQL version" flag when automagically building our mapping list of internal PG data types to SQL types. This was causing a bug in which a SQL_CHAR type was getting mapped to a pg "char" (single character) rather than the expected bpchar (both map back to SQL_CHAR). This was causing bind_param to fail, if it was using the returned types. Per email to dbi-dev list from H.Merijn Brand 2013/03/04.
Add a test for the recent SQL_CHAR related fix.
Note recent change.
Adjust test count
Make sure this test matchess the description!
Credit recent patch
Add TODO (aka known failing) tests for setting $sth->err to 6 as specified in the docs.
Yes, it's all utf-8 now.
Minor doc changes for utf8 stuff.
Cleanup Changes
Note recent Changes
Correct the number of .pm files we skip for this test.
Fix skip numbers
Bump version to DBI to 1.54 everywhere. Credit recent change.
Bump Perl requirement to 5.8.4. May go up further, but let's put this stake in the ground.
Make some tests AUTHOR_TESTING, not RELEASE_TESTING, as we really don't want anyone else to run them.
Adjust test to use 3 digit exponential values (CPAN bug #59449)
gitignore tweaks
Move Perl requirement down to 5.8.1, per RT request.
Quick dump of lo_truncate support. Not working yet.
Let's call this TODO item done.
Add a new field, pg_expression, to the results from statistics_info. This
Note latest change
Make sure we force expand_array on inside of foreign_key_info, then reset it back the way it was.
Skip the still-not-working pg_lo_truncate tests.
Stop trying to count tests here!
Update with latest type information from Postgres
Bump Test::More requirement to .88, as we are using done_testing()
Bump version here too
Spellcheck tweaks.
Perl::Critic tweaks
Testing tweaks
Bump DBI requirement to 1.614.
Better git ignoring.
Cleanup the Changes file
Adjust debugging to use new _slow macros.
Skip pg_lo_import_with_oid tests for now. Noted as RT #90448
Teach everyone that AutoInactiveDestroy is a legal word for us.
pg_server_prepare should be 1 here. So this test is not working yet!
More lo_import_with_oid tweaks. May just yank the whole thing out! :)
Minor Changes tweaks
Version 2.20.1_2
Minor casting tweak
Switch from char to SV, per ideas from the coropg project.
Remove dead link
Add new attribute, pg_switch_prepared, which controls when we switch
Note TODO item
Slight adjustment to new switch_prepared logic.
Add tests and attribute handlers for $h->{switch_prepared}
Note some recent changes inside Changes
Update TODO
Spellcheck tweaks
Use newer escaping syntax for pod
Version 2.20.1_3
do() blocks causing errors in cpantesters: perhaps the use lib is being ignored inside of it. Changed to a simple if, and added a test that we have a sane POSTGRES_HOME and bail at that point if we do not.
Version 2.20.1_4
Duh. Problem wasn't in Makefile.PL, it was failure to add a new file to the MANIFEST.
Set to version 2.20.1_5
Make arrayout field a hair larger to account for growing size in PG source. Merely pretties up types.c
Make sure we rewrite the new done_testing bit
Add recent file to README.dev
Internal name should be INT8, not LONGINT
Minor tweaks to testing script
SvUTF8_on is NOT enough here. Total testing fail: should be catching this.
Add some UTF-8 tests. Fix typo in dbdimp.c
More sv utf8 tweaks
Tweak test due to new utf8
Note recent Change
Note latest patch
Add utf-8 test from cpan bug 91655
More UTF-8 tweaking; tweak tests to use the new 'use charnames' pragma. Use Encode to strip utf8 flag for testing.
This test works again
Bump to version 2.20.1_6
Credit last patch
Add support to start testing arrays of bytea
Don't flip the utf8 flag on when we are unrolling arrays of bytea.
Respect pg_bool_tf when unrolling arrays.
Try to dequote bytea array, per bug 91454. Untested!
Version bump to 3.0.0, unless and until we make more changes.
Test cleanup
Note recent bytea[] change
Remove debugging message.
Oops, things are coming back here as t/f
Thinko. Thank goodness for tests
Add a new env DBDPG_NOCLEANUP for debugging. Prevents t/99cleanup.t from doing anything.
Fix test counting, should fix RT #79544.
Note latest change
Ignores
Change debugging env PGINITDB to DBDPG_INITDB.
Rainer Weikusat (1):
Introduce a _fatal_sqlstate function setting the sqlstate of the dbh either to
Spencer Sun (1):
Format our prepared statements with %x, not %d, for very long-running transactions in which the internal prepare_number rolls over into negative numbers - which bumps up against Postgres' rules about prepared statement names. Per RT #88827.
Stephen Keller (1):
Better handling of failed libpq connections. Should fix the bug where a SQLSTATE of 08000 is expected but not returned.
-----------------------------------------------------------------------
No new revisions were added by this update.
--
Alioth's /usr/local/bin/git-commit-notice on /srv/git.debian.org/git/pkg-perl/packages/libdbd-pg-perl.git
More information about the Pkg-perl-cvs-commits
mailing list