Getting everything ready for release of 1.0019_001
[dbsrgits/DBM-Deep.git] / Changes
CommitLineData
ffed8b01 1Revision history for DBM::Deep.
2
67e9b86f 31.0019_001 Dec 31 22:00:00 2009 EST
4 (This is the first developer release for 1.0020.)
5 (This version is compatible with 1.0014)
6 - DBM::Deep has been refactored to allow for multiple engines. There are two
7 engines built so far:
8 - File (the original engine)
9 - DBI (an engine based on DBI)
10 - The DBI engine has only been tested on MySQL and isn't transactional.
11 - InnoDB sucks horribly. When run in a sufficient isolation mode, it
12 creates deadlocks.
13 - A custom Build.PL has been written to allow for running tests under
14 CPAN.pm against the various engines.
15 - This also allows running the long tests under CPAN.pm
16 - This has meant a ton of refactoring. Hopefullly, this refactoring will
17 allow finding some of the niggly bugs more easily. Those tests have not
18 been enabled yet. That's the next developer release.
19 - Hopefully, this multi-engine support will allow deprecation of the file
20 format in the future.
4f0f6fff 21
576320ff 221.0014 Jun 13 23:15:00 2008 EST
23 - (This version is compatible with 1.0013)
24 - Fix for RT#36781 (t/44 has an unrequired dependency)
5c0756fc 25 - lock() has been aliased to lock_exclusive(). There is now
26 a lock_shared() method. The :flock constants are no longer
27 imported into the DBM::Deep namespace.
28 **** THIS IS AN API CHANGE ****
576320ff 29
9d3b38e4 301.0013 Jun 13 23:15:00 2008 EST
695c88b1 31 - (This version is compatible with 1.0012)
32 - Fix for RT#30144 (Optimization failure on Win32)
33 - Fixed a bug in reindex_entry (Thanks, Wulfram Humann!)
34
08164b50 351.0012 Jun 09 15:00:00 2008 EST
36 - (This version is compatible with 1.0011)
37 - Fix for RT#30085 (Remove dependency on XS module)
38 - Thank you very much tachyon-II@Perlmonks!!
9d3b38e4 39 - This also resolves RT#35424 (DBM::Deep breaks XML::Twig).
08164b50 40 - Updated the POD with fixes that were made, but still on the TODO list.
41 - Bypass for RT#36419 (t/44_upgrade_db.t fails on *BSD)
42 - We're just going to skip that for now. It's a number-of-processes-open
43 issue. Best is to do a port to another module to do the heavy lifting.
44
451.0011 May 27 15:00:00 2008 EST
46 - (This version is compatible with 1.0010)
47 - A test has the wrong plan.
48
edd45134 491.0010 May 27 12:00:00 2008 EST
b8370759 50 - (This version is compatible with 1.0009)
51 - Fix for RT#35140 (invalid POD links)
edd45134 52 - Fix for RT#34819 (Cannot assign the same value back to the same location)
53 - Fix for RT#29957 (Cannot assign the same value back to the same location)
54 - Fix for RT#33863 (Cannot shift an arrayref from an array)
55 - When something is deleted from a DB, the value is export()ed, allowing it
56 to be saved.
57 - This exporting is only done if the refcount == 0 after the deletion.
b8370759 58
6e6789b0 591.0009 Mar 19 12:00:00 2008 EDT
60 - (This version is compatible with 1.0008)
61 - Internal refactorings to prepare for some optimizations.
62 - _fh() has been removed. It was marked as private, so don't complain.
63 - Skip a test that was spuriously failing on Win32 (Thanks, Alias!)
64
45f047f8 651.0008 Mar 09 20:00:00 2008 EDT
66 - (This version is compatible with 1.0007)
67 - Fixed a number of Win32 issues (Reported by Steven Samelson - thank you!)
68 - Much thanks to Nigel Sandever and David Golden for their help
69 debugging the issues, particularly with DBM::Deep's usage of
70 File::Temp (which removes a number of warnings).
71 - Autovivification now works on Win32. It turns out that when a
72 process takes a shared flock on a file, it's not allowed to write to
73 it under Win32, unlike *nix. This is probably a good catch.
74 - Note: The fix is a hack. All locks are now exclusive until a
75 better fix is found.
76
5a70a6c0 771.0007 Jan 10 00:00:00 2008 EDT
78 - (This version is compatible with 1.0006)
79 - Applied a patch+failing test submitted by sprout@cpan.org. Thanks!
80 - Turns out that the case of 17 keys with the same first character in the
81 MD5 hash wasn't being tested for. This was a crashbug.
82 - A fix has been made to upgrade_db.pl (RT# 30067)
83 - The version determinations were in the wrong order or evaluation. This
84 meant that upgrade_db.pl wouldn't work as expected (or at all).
85 - Added a minimum Pod::Usage requirement (RT# 29976)
86 - It's an optional item in Build.PL
87 - utils/upgrade_db.pl now checks for that version, as does the test.
88
e00d0eb3 891.0006 Oct 01 23:15:00 2007 EDT
90 - (This version is compatible with 1.0005)
91 - Removed Clone and replaced it with a hand-written datastructure walker.
92 - This greatly reduces the footprint of a large import
93 - This bypasses a failure of Clone under Perl 5.9.5
94 - Moved t/37_delete_edge_cases.t to t_attic because it wasn't really used
95 - import() has a stricter API now. This is a potentially incompatible API
96 change. Only HASH and ARRAY refs are now allowed and they must match the type
97 of the object being imported into.
98
c57b19c6 991.0005 Oct 01 11:15:00 2007 EDT
100 - (This version is compatible with 1.0004)
101 - Added proper singleton support. This means that the following now works:
102 $db->{foo} = [ 1 .. 3];
103 my $x = $db->{foo};
104 my $y = $db->{foo};
105 is( $x, $y ); # Now passes
106 - This means that Data::Dumper now properly reports when $db->{foo} = $db->{bar}
107
888453b9 1081.0004 Sep 28 12:15:00 2007 EDT
109 - (This version is compatible with 1.0003)
110 - Fixed the Changes file (wrong version was displayed for 1.0003)
111 - Added filter sugar methods to be more API-compatible with other DBMs
112 - This was added to support a patch provided to IO::All so it can
113 use DBM::Deep as a DBM provider.
114 - Implemented _dump_file in order to display the file structure. As a
115 result, the following bugs were fixed:
116 - Arrays and hashes now clean up after themselves better.
117 - Bucketlists now clean up after themselves better.
118 - Reindexing properly clears the old bucketlist before freeing it.
119
1201.0003 Sep 24 14:00:00 2007 EDT
1cff45d7 121 - THIS VERSION IS INCOMPATIBLE WITH FILES FROM ALL OTHER PRIOR VERSIONS.
122 - Further fixes for unshift/shift/splice and references (RT# 29583)
123 - To fix that, I had to put support for real references in.
124 - the 16 and 22 tests are now re-enabled.
125 - Yes, this means that real references work. See t/45_references.t
126
807f63a7 1271.0002 Sep 20 22:00:00 2007 EDT
128 - (This version is compatible with 1.0001)
129 - Expanded _throw_error() so that it provides better information.
130 (Thanks brian d foy!)
131 - Fixed how shift, unshift, and splice work when there are references
132 being moved. It now no longer dies.
133 - Added diag in t/17_import.t to note that the failing tests on blead
134 are due to Clone being broken, not DBM::Deep. The tests will still
135 fail because I don't want users to install something that's broken
136 and deal with those bug reports.
137
151e0077 1381.0001 Mar 12 16:15:00 2007 EDT
139 - (This version is compatible with 1.0000)
140 - Added a missing dependency on IO::Scalar (RT #25387)
141 - Fixed how t/44_upgrade_db.t and utils/upgrade_db.pl worked
142 - utils/upgrade_db.pl now uses #!perl, not #!/usr/bin/perl
143 - t/44_upgrade_db.t now explicitly calls $^X
144 (Thanks merlyn!)
145
1461.0000 Feb 26 22:30:00 2007 EST
e9b0b5f0 147 - THIS VERSION IS INCOMPATIBLE WITH FILES FROM ALL OTHER PRIOR VERSIONS.
148 - To aid in this form of upgrades, DBM::Deep now checks the file format
149 version to make sure that it knows how to read it.
150 - db_upgrade.pl was added to utils/. This will -NOT- install onto
151 your system. This is deliberate.
152 - db_upgrade.pl will not handle developer release file formats. This
153 is due to the fact that all developer releases in preparation for a
154 given release share the same file version, even though the file
155 format may change. This is deliberate.
156 - Importing no longer takes place within a transaction
157 - The following parameters were added:
158 - data_sector_size - this determines the default size of a data sector.
159 - Correctly handle opening readonly files
160
151e0077 1610.99_04 Jan 24 22:30:00 2007 EST
2120a181 162 - Added the missing lib/DBM/Deep.pod file to the MANIFEST
163 - Fixed a poorly-designed test that was failing depending on what Clone::Any
e9b0b5f0 164 was using.
2120a181 165 - All "use 5.6.0;" lines are now "use 5.006_000;" to avoid warnings about
e9b0b5f0 166 unsupported vstrings in 5.9.x
2120a181 167
151e0077 1680.99_03 Jan 23 22:30:00 2007 EST
2120a181 169 - THIS VERSION IS INCOMPATIBLE WITH FILES FROM ALL OTHER PRIOR VERSIONS.
170 - The fileformat changed completely. I will be writing a converter, but
171 it's not there right now. Do NOT expect that this module will
172 correctly detect older versions and handle them sanely. Sanity will be
173 there for 1.00, but we're not there yet, are we?
a8fdabda 174 - Converted to use FileHandle::Fmode to handle filehandle status checks
13ff93d5 175 - Fixed bug with deleting already-deleted items on Win32 (reported by Nigel Sandever)
2120a181 176 - The guts of how transactions work has been rewritten to better handle
177 some edgecases. This required a complete rewrite of the engine.
178 - Freespace management is now in place. It's not perfect, but it's there.
179 - The rewrite of the engine required a rewrite of how first_key/next_key
180 was implemented. This should result in significant speed improvements.
181 - Self-reference has been removed. This means you cannot do:
182 $db->{foo} = { x => 'y' };
183 $db->{bar} = $db->{foo};
184 I hope to be able to return this functionality by 1.00, but I cannot
185 promise anything. To do this properly, it requires refcounting in order
186 to correctly handle deletions and transactions. Once you move away from
187 a simple tree, everything becomes really hard.
a8fdabda 188
c8602313 1890.99_02 Apr 28 05:00:00 2006 Pacific
190 - Added missing file to the MANIFEST
191
1920.99_01 Apr 27 18:00:00 2006 Pacific
d2aefba0 193 - Added explicit dependency on Perl 5.6.0
e06824f8 194 - Digest::MD5 requires 5.6.0
195 - Sub::Uplevel (dep of Test::Exception) requires 5.6.0
95967a5e 196 - Removed error()/clear_error()
d2aefba0 197 - All error-handling is done with die()
e06824f8 198 - Broke out DBM::Deep's code into DBM::Deep::Engine
d2aefba0 199 - Tied variables can no longer be assigned to a DBM::Deep object.
200 - This includes cross-file assignments.
201 - Autovivification now works
202 - This is a consequence of the fact that all assignments are tied.
203 - set_pack() and set_digest() have been removed.
204 - Instead, you will now pass the appropriate values into new()
205 - A pack_size parameter has been added to make 64-bit files easier
c8602313 206 - Transactions now work
95967a5e 207
a8fdabda 2080.983 Apr 10 20:00:00 2006 Pacific
209 - Added patch inspired by Jeff Janes (Thanks!)
210 - Autovivification now works correctly
211 - The following now works correctly
212 my %hash = ( a => 1 );
213 $db->{hash} = \%hash;
214 $hash{b} = 2;
215 cmp_ok( $db->{hash}{b}, '==', 2 );
216 - (RT#18530) - DBM::Deep now plays nicely with -l
217
2180.982 Mar 08 11:00:00 2006 Pacific
219 - Fixed smoketests that were failing on Win32
220 - Added restriction for Perl 5.6.0 or higher.
221 - Digest::MD5 and Sub::Uplevel (dep of Test::Exception) require 5.6+
222
2230.981 Mar 06 11:00:00 2006 Pacific
224 - (RT#17947) - Fixed test that was failing on older Perls
225
9be51a89 2260.98 Feb 28 11:00:00 2006 Pacific
a8bd842e 227 - Added in patch by David Cantrell to allow use of DATA filehandle
bcbe5355 228 - Fixed bug where attempting to export() a structure that used autobless would die
e3cb84dc 229 - Fixed arraytest slowness by localizing $SIG{__DIE__} to prevent Test::Builder's
230 $SIG{__DIE__} from being called. (q.v. http://perldoc.perl.org/functions/eval.html)
4d35d856 231 - More methods have been made private:
232 - root() is now _root()
233 - base_offset() is now _base_offset()
234 - fh() is now _fh()
235 - type() is now _type()
236 - precalc_sizes() is now _precalc_sizes()
e3cb84dc 237
6d049f2c 2380.97 Feb 24 10:00:00 2006 Pacific
e096f245 239 - Reorganization of distribution to a more standard layout
ffed8b01 240 - Migration to Module::Build with EU::MM backwards compatibility
e096f245 241 - Migration of all tests to use Test::More and Test::Exception
242 - Added Devel::Cover report to DBM::Deep POD
243 - Test coverage improved to 89.6% (and climbing)
8b957036 244 - The following methods have been renamed to reflect their private nature:
245 - init() is now _init()
14a3acb6 246 - open() is now _open()
0207d648 247 - close() is now _close()
261d1296 248 - load_tag() is now _load_tag()
249 - index_lookup() is now _index_lookup()
250 - add_bucket() is now _add_bucket()
251 - get_bucket_value() is now _get_bucket_value()
252 - delete_bucket() is now _delete_bucket()
253 - bucket_exists() is now _bucket_exists()
254 - find_bucket_list() is now _find_bucket_list()
255 - traverse_index() is now _traverse_index()
256 - get_next_key() is now _get_next_key()
257 - copy_node() is now _copy_node()
258 - throw_error() is now _throw_error()
217fef02 259 - The various tied classes have been broken out. This means that testing
260 "ref( $obj ) eq 'DBM::Deep'" will now fail. The correct test is
261 "eval { $obj->isa( 'DBM::Deep' ) }".
262 - The various methods like push and delete now have the same return values as
263 the standard builtins.
9d085683 264 - TIEARRAY and TIEHASH now check their parameters more thoroughly
e096f245 265 - Negative indices for arrays works as expected, including throwing the appropriate
266 errors.
26138c00 267 - RT #16877 is fixed (DBM::Deep broken with Perl 5.9.3+).
268 - RT #14893 is fixed (tie() and new() use different parameter lists).
269 - A bug with optimize and threading is fixed.
270 - autobless has received some attention, resulting in a number of bugs fixed.
90f93b43 271 - Removed mode option as it just caused confusion.
3ea61317 272 - Removed volatile option as it is pretty useless (use locking instead)
273 - Locking now implicitly enables autoflush
ffed8b01 274
2750.96 Oct 14 09:55:00 2005 Pacific
276 - Fixed build (OS X hidden files killed it)
277 - You can now pass in an optional filehandle to the constructor
278
2790.95 Oct 12 13:58:00 2005 Pacific
280 - Added optional autobless flag to preserve and restore blessed hashes
281 - Fixed bug where 0 could not be fetched using get_next_key
282 - Fixed bug where tie() constructor didn't accept a hash ref for args
283 - optimize() now preserves user/group/permissions
284 - Errors are now FATAL (meaning it calls die()), unless you set debug flag
285
2860.94 Apr 13 19:00:26 2004 Pacific
287 - Fixed bug reported by John Cardenas (corruption at key level when
288 replace of less data was done on bucket)
289
2900.93 Feb 15 19:53:17 2004 Pacific
291 - Fixed optmize() on Win32 where orig file couldn't be overwritten unless
292 filehandle was closed first. This change introduces a potential race
293 condition when using locking and optmize() on Win32, but it can be
294 fixed in the future using a soft copy instead of Perl's rename().
295
2960.92 Feb 12 19:10:22 2004 Pacific
297 - Fixed bug where passing a reference to a different DBM::Deep object
298 would still result in an internal reference.
299 - Added export() method for recursively extracting hashes/arrays into
300 standard in-memory Perl structures.
301 - Added import() method for recursively importing existing Perl hash/
302 array structures
303 - Fixed bug where optimize() wouldn't work if base level of DB was
304 an array instead of a hash.
305
3060.91 Feb 12 02:30:22 2004 Pacific
307 - Fixed bug with splice() when length of removed section was 0
308 - Updated POD re: circular refs and optimize()
309 - Had to jump version numbers to 0.91 because in previous releases
310 I used only a single digit after the decimal which was confusing
311 the CPAN indexer.
312
3130.10 Feb 11 08:58:35 2004 Pacific
314 - Fixed bug where default file mode was CLEARING files (Thanks Rich!)
315 - Added experimental support for circular references
316 - Fixed bugs in shift(), unshift() and splice() where nested objects
317 in array would be recursively re-stored as basic hashes/arrays
318 - Fixed typos in POD docs
319
3200.9 Feb 10 03:25:48 2004 Pacific
321 - Added Filters for storing/fetching keys/values
322 - Added hook for supplying own hashing algorithm
323 - FIxed some typos in POD docs, added new sections
324
3250.8 Feb 8 02:38:22 2004 Pacific
326 - Renamed to DBM::Deep for CPAN
327 - Added optimize() method for rekindling unused space
328 - Now returning hybrid tie()/OO object from new()
329 - Basic error handling introduced
330 - Added debug mode for printing errors to STDERR
331 - Added TYPE_HASH and TYPE_ARRAY constants for "type" param
332 - Added clone() method for safe copying of objects
333 - Wrote POD documentation
334 - Added set_pack() function for manipulating LONG_SIZE / LONG_PACK
335 - Added aliases for most tied functions for public use
336 - Now setting binmode() on FileHandle for Win32
337 - Added 45 unit tests
338
3390.7 Jan 4 11:31:50 2003 UTC
340 - Renamed to DeepDB
341 - Changed file signature to DPDB (not compatible with older versions)
342 - Converted array length to packed long instead of sprintf()ed string
343
3440.6 Dec 31 15:12:03 2002 UTC
345 - Some misc optimizations for speed
346
3470.5 Oct 18 08:55:29 2002 UTC
348 - support for force_return_next parameter in traverse_index() method for
349 ultra-fast combined key search/removal
350
3510.4 Oct 15 20:07:47 2002 UTC
352 - now making sure filehandle is open for all DB calls
353
3540.3 Oct 3 19:04:13 2002 UTC
355 - fixed bug that could cause corrupted data when using locking
356
3570.2 Aug 6 16:37:32 2002 UTC
358 - Removed base index caching, as it can cause problems when two processes
359 are populating the db at the same time (even with locking)
360
3610.1 Jun 3 08:06:26 2002 UTC
362 - initial release