pg 0.17.1 → 0.18.4
Sign up to get free protection for your applications and to get access to all the features.
- checksums.yaml +4 -4
- checksums.yaml.gz.sig +0 -0
- data/ChangeLog +2407 -2
- data/History.rdoc +68 -0
- data/Manifest.txt +29 -1
- data/README-Windows.rdoc +15 -26
- data/README.rdoc +52 -2
- data/Rakefile +56 -18
- data/Rakefile.cross +77 -49
- data/ext/extconf.rb +33 -26
- data/ext/pg.c +142 -21
- data/ext/pg.h +242 -6
- data/ext/pg_binary_decoder.c +162 -0
- data/ext/pg_binary_encoder.c +162 -0
- data/ext/pg_coder.c +479 -0
- data/ext/pg_connection.c +858 -553
- data/ext/pg_copy_coder.c +561 -0
- data/ext/pg_errors.c +6 -0
- data/ext/pg_result.c +479 -128
- data/ext/pg_text_decoder.c +421 -0
- data/ext/pg_text_encoder.c +663 -0
- data/ext/pg_type_map.c +159 -0
- data/ext/pg_type_map_all_strings.c +116 -0
- data/ext/pg_type_map_by_class.c +239 -0
- data/ext/pg_type_map_by_column.c +312 -0
- data/ext/pg_type_map_by_mri_type.c +284 -0
- data/ext/pg_type_map_by_oid.c +355 -0
- data/ext/pg_type_map_in_ruby.c +299 -0
- data/ext/util.c +149 -0
- data/ext/util.h +65 -0
- data/lib/pg/basic_type_mapping.rb +399 -0
- data/lib/pg/coder.rb +83 -0
- data/lib/pg/connection.rb +81 -29
- data/lib/pg/result.rb +13 -3
- data/lib/pg/text_decoder.rb +44 -0
- data/lib/pg/text_encoder.rb +27 -0
- data/lib/pg/type_map_by_column.rb +15 -0
- data/lib/pg.rb +12 -2
- data/spec/{lib/helpers.rb → helpers.rb} +101 -39
- data/spec/pg/basic_type_mapping_spec.rb +251 -0
- data/spec/pg/connection_spec.rb +516 -218
- data/spec/pg/result_spec.rb +216 -112
- data/spec/pg/type_map_by_class_spec.rb +138 -0
- data/spec/pg/type_map_by_column_spec.rb +222 -0
- data/spec/pg/type_map_by_mri_type_spec.rb +136 -0
- data/spec/pg/type_map_by_oid_spec.rb +149 -0
- data/spec/pg/type_map_in_ruby_spec.rb +164 -0
- data/spec/pg/type_map_spec.rb +22 -0
- data/spec/pg/type_spec.rb +697 -0
- data/spec/pg_spec.rb +24 -18
- data.tar.gz.sig +0 -0
- metadata +111 -45
- metadata.gz.sig +0 -0
data/ChangeLog
CHANGED
@@ -1,8 +1,2413 @@
|
|
1
|
+
2015-11-13 Michael Granger <ged@FaerieMUD.org>
|
2
|
+
|
3
|
+
* .hgignore:
|
4
|
+
Ignore some build artifacts
|
5
|
+
[f4bd9e53ccab] [tip]
|
6
|
+
|
7
|
+
* Rakefile:
|
8
|
+
Only use fivefish formatter when building from repo
|
9
|
+
[6161d8eeb3f1]
|
10
|
+
|
11
|
+
2015-11-13 Lars Kanis <lars@greiz-reinsdorf.de>
|
12
|
+
|
13
|
+
* History.rdoc:
|
14
|
+
Update History.rdoc regarding PR#22.
|
15
|
+
[e4a8a1d41104]
|
16
|
+
|
17
|
+
* Merge ../ruby-pg.git.bare
|
18
|
+
[810c1e4dad4d]
|
19
|
+
|
20
|
+
2015-10-08 Lars Kanis <larskanis@gmail.com>
|
21
|
+
|
22
|
+
* README.rdoc:
|
23
|
+
Fix link to appveyor CI tests in the README.
|
24
|
+
[604f4dd3d5bd]
|
25
|
+
|
26
|
+
2015-11-04 Lars Kanis <lars@greiz-reinsdorf.de>
|
27
|
+
|
28
|
+
* ext/pg_binary_encoder.c:
|
29
|
+
Merged in t-richards/ruby-pg/jemalloc-native-ext (pull request #22)
|
30
|
+
|
31
|
+
Fix build against jemalloc 4.0.0
|
32
|
+
[72168db5fb7b]
|
33
|
+
|
34
|
+
2015-09-10 Tom Richards <tom@tomrichards.net>
|
35
|
+
|
36
|
+
* ext/pg_binary_encoder.c:
|
37
|
+
Fix build against jemalloc 4.0.0
|
38
|
+
[fe05106909f7] <jemalloc-native-ext>
|
39
|
+
|
40
|
+
2015-09-24 Lars Kanis <lars@greiz-reinsdorf.de>
|
41
|
+
|
42
|
+
* History.rdoc:
|
43
|
+
Add latest changes to History.rdoc.
|
44
|
+
[11b978b08d63]
|
45
|
+
|
46
|
+
* History.rdoc:
|
47
|
+
Add missing entry in History.rdoc.
|
48
|
+
|
49
|
+
This fixes bitbucket issue #225 : https://bitbucket.org/ged/ruby-
|
50
|
+
pg/issues/225
|
51
|
+
[132d674d999f]
|
52
|
+
|
53
|
+
* Merge ../ruby-pg.git.bare
|
54
|
+
[479799a98cfa]
|
55
|
+
|
56
|
+
* ext/pg_text_encoder.c, spec/pg/connection_spec.rb,
|
57
|
+
spec/pg/type_spec.rb:
|
58
|
+
Avoid segfault, when quote_ident is called with Array containing
|
59
|
+
non-strings.
|
60
|
+
|
61
|
+
This fixes bitbucket issue #226 : https://bitbucket.org/ged/ruby-
|
62
|
+
pg/issues/226
|
63
|
+
[4f0bd19e9111]
|
64
|
+
|
65
|
+
2015-09-08 Lars Kanis <lars@greiz-reinsdorf.de>
|
66
|
+
|
67
|
+
* Merge pull request #10 from PuzzleFlow/master
|
68
|
+
|
69
|
+
Fixing compilation problems with Microsoft Visual Studio 2008.
|
70
|
+
[dfb62954f2d6]
|
71
|
+
|
72
|
+
2015-09-08 jfali <jfali@puzzleflow.com>
|
73
|
+
|
74
|
+
* ext/extconf.rb, ext/pg_binary_decoder.c, ext/pg_binary_encoder.c,
|
75
|
+
ext/pg_text_decoder.c, ext/pg_text_encoder.c, ext/pg_type_map.c,
|
76
|
+
ext/pg_type_map_by_column.c:
|
77
|
+
Fixing compilation problems with Microsoft Visual Studio 2008.
|
78
|
+
- inttypes.h is not available in MSVC 2008
|
79
|
+
- Structs members must be initialized implicitly
|
80
|
+
[fcf731d3dff7]
|
81
|
+
|
82
|
+
2015-09-07 Lars Kanis <lars@greiz-reinsdorf.de>
|
83
|
+
|
84
|
+
* README.rdoc:
|
85
|
+
Merge pull request #8 from gitter-badger/gitter-badge
|
86
|
+
|
87
|
+
Add a Gitter chat badge to README.rdoc
|
88
|
+
[ab4d5656abce]
|
89
|
+
|
90
|
+
2015-03-09 The Gitter Badger <badger@gitter.im>
|
91
|
+
|
92
|
+
* README.rdoc:
|
93
|
+
Added Gitter badge
|
94
|
+
[beb02a90f948]
|
95
|
+
|
96
|
+
2015-09-03 Michael Granger <ged@FaerieMUD.org>
|
97
|
+
|
98
|
+
* .hgtags:
|
99
|
+
Added tag v0.18.3 for changeset 01c42c68797e
|
100
|
+
[4c8df76aff6f]
|
101
|
+
|
102
|
+
* .hgsigs:
|
103
|
+
Added signature for changeset 57291f1e96b9
|
104
|
+
[01c42c68797e] [v0.18.3]
|
105
|
+
|
106
|
+
* .hoerc:
|
107
|
+
Exclude some generated files from the Manifest
|
108
|
+
[57291f1e96b9]
|
109
|
+
|
110
|
+
* History.rdoc, lib/pg.rb:
|
111
|
+
Bump patch version, update history
|
112
|
+
[3b49541c9f12]
|
113
|
+
|
114
|
+
2015-09-01 Lars Kanis <lars@greiz-reinsdorf.de>
|
115
|
+
|
116
|
+
* lib/pg/text_decoder.rb:
|
117
|
+
Add missing require
|
118
|
+
[d0ba6a2cb10f]
|
119
|
+
|
120
|
+
* Merge pull request #9 from ramfjord/date-decoder-return-dates
|
121
|
+
|
122
|
+
Date decoder to return real Dates
|
123
|
+
[676def89b826]
|
124
|
+
|
125
|
+
2015-06-21 Thomas Ramfjord <thomas.ramfjord@gmail.com>
|
126
|
+
|
127
|
+
* lib/pg/text_decoder.rb, spec/pg/basic_type_mapping_spec.rb:
|
128
|
+
modified TextDecoder::Date to return an actual ruby Date instead of
|
129
|
+
a Time
|
130
|
+
[c50593d12ba2]
|
131
|
+
|
132
|
+
2015-08-20 Lars Kanis <lars@greiz-reinsdorf.de>
|
133
|
+
|
134
|
+
* Rakefile:
|
135
|
+
Add include for Hoe::MercurialHelpers as a temporary solution for
|
136
|
+
the error:
|
137
|
+
|
138
|
+
Packaging tasks require the hoe-mercurial plugin (gem install hoe-
|
139
|
+
mercurial)
|
140
|
+
[a84f83ff5b31]
|
141
|
+
|
142
|
+
* History.rdoc:
|
143
|
+
Update History.rdoc
|
144
|
+
[cadfac537e73]
|
145
|
+
|
146
|
+
* Merge branch 'master' of github.com:larskanis/ruby-pg
|
147
|
+
[d27d6d5a6545]
|
148
|
+
|
149
|
+
* README.rdoc:
|
150
|
+
Add Appveyor status badge to the README.
|
151
|
+
[c9637c6efcfb]
|
152
|
+
|
153
|
+
* appveyor.yml:
|
154
|
+
Use 9.3 because 9.0, 9.1 and 9.2 segfault in PG.init_openssl
|
155
|
+
[158f6649b7b8]
|
156
|
+
|
157
|
+
* spec/pg/connection_spec.rb:
|
158
|
+
Skip conn#host test on Windows-x64, because it segfaults, currently.
|
159
|
+
[04110aad7fc6]
|
160
|
+
|
161
|
+
* spec/pg/connection_spec.rb:
|
162
|
+
libpq on Windows doesn't respect the PGPORT env variable, so we
|
163
|
+
accept both.
|
164
|
+
[b4b666011f3f]
|
165
|
+
|
166
|
+
* spec/helpers.rb:
|
167
|
+
Older versions of PostgreSQL don't have a pid column, so this could
|
168
|
+
raise an exception: TypeError: can't convert nil into Integer
|
169
|
+
[48f1145fa631]
|
170
|
+
|
171
|
+
* Rakefile.cross:
|
172
|
+
Update OpenSSL version for Windows cross build.
|
173
|
+
[be6b75fef0ff]
|
174
|
+
|
175
|
+
* appveyor.yml:
|
176
|
+
Add CI-tests on appveyor.
|
177
|
+
[c2dc9e491853]
|
178
|
+
|
179
|
+
2015-06-29 Lars Kanis <kanis@comcard.de>
|
180
|
+
|
181
|
+
* History.rdoc:
|
182
|
+
Merge branch 'master' of https://github.com/larskanis/ruby-pg
|
183
|
+
|
184
|
+
Conflicts: History.rdoc
|
185
|
+
[affb47533d82]
|
186
|
+
|
187
|
+
* History.rdoc, ext/util.c, spec/pg/type_spec.rb:
|
188
|
+
Fix data type resulting in wrong base64 encoding.
|
189
|
+
[5fb9170f6a7d]
|
190
|
+
|
191
|
+
2015-06-17 Lars Kanis <lars@greiz-reinsdorf.de>
|
192
|
+
|
193
|
+
* Gemfile, Rakefile, Rakefile.cross:
|
194
|
+
Update versions of rake-compiler-dock, OpenSSL and PostgreSQL.
|
195
|
+
[10c6b496314c]
|
196
|
+
|
197
|
+
2015-06-12 Lars Kanis <lars@greiz-reinsdorf.de>
|
198
|
+
|
199
|
+
* Rakefile.cross:
|
200
|
+
Update OpenSSL and PostgreSQL versions for cross build.
|
201
|
+
[02094cca1cf8]
|
202
|
+
|
203
|
+
2015-06-11 Lars Kanis <lars@greiz-reinsdorf.de>
|
204
|
+
|
205
|
+
* History.rdoc, README-Windows.rdoc:
|
206
|
+
Update README-Windows in regards to rake-compiler-dock.
|
207
|
+
[4d9fd821ec1e]
|
208
|
+
|
209
|
+
2015-06-09 Lars Kanis <kanis@comcard.de>
|
210
|
+
|
211
|
+
* Rakefile.cross:
|
212
|
+
Cross build: Ensure ChangeLog file is built on the host side before
|
213
|
+
entering the docker container.
|
214
|
+
[d32412acff38]
|
215
|
+
|
216
|
+
* Rakefile.cross:
|
217
|
+
Cross build: Add workaround for missing symlink support on Windows.
|
218
|
+
[52bb3eef00f0]
|
219
|
+
|
220
|
+
* Rakefile.cross:
|
221
|
+
Explicitly set the gcc-toolchain, instead of deriving from the rake-
|
222
|
+
compiler config.
|
223
|
+
|
224
|
+
Now, that we use the libpq.dll (instead of static linking), we can
|
225
|
+
mix the gcc-toolchains. So we better build the dll with the newer
|
226
|
+
mingw-w64 toolchain (instead of the old mingw32) and build only the
|
227
|
+
extension for ruby-1.9 with the old mingw32-gcc.
|
228
|
+
[b3d169fa189b]
|
229
|
+
|
230
|
+
* Merge https://github.com/ged/ruby-pg
|
231
|
+
[5aeb99dda0ca]
|
232
|
+
|
233
|
+
2015-05-25 Michael Granger <ged@FaerieMUD.org>
|
234
|
+
|
235
|
+
* lib/pg/basic_type_mapping.rb:
|
236
|
+
Whitespace fixes
|
237
|
+
[57cfbc230597]
|
238
|
+
|
239
|
+
2015-05-14 Michael Granger <ged@FaerieMUD.org>
|
240
|
+
|
241
|
+
* .hgtags:
|
242
|
+
Added tag v0.18.2 for changeset 22a361201fd1
|
243
|
+
[00d6997551e6]
|
244
|
+
|
245
|
+
* .hgsigs:
|
246
|
+
Added signature for changeset 7d31b04e7913
|
247
|
+
[d77ab8ab870e]
|
248
|
+
|
249
|
+
* History.rdoc, lib/pg.rb:
|
250
|
+
Bump patch version, update History
|
251
|
+
[fa56690fe72d]
|
252
|
+
|
253
|
+
* Merged with git branch.
|
254
|
+
[9b2c80df821d]
|
255
|
+
|
256
|
+
2015-06-09 Lars Kanis <kanis@comcard.de>
|
257
|
+
|
258
|
+
* Rakefile.cross:
|
259
|
+
Cross build: Copy gem signing key and certs to be accessable from
|
260
|
+
the docker container
|
261
|
+
[c5382a785f4d]
|
262
|
+
|
263
|
+
2015-06-08 Lars Kanis <lars@greiz-reinsdorf.de>
|
264
|
+
|
265
|
+
* Merged
|
266
|
+
[3d56a65c78c5]
|
267
|
+
|
268
|
+
* Gemfile, Rakefile, Rakefile.cross:
|
269
|
+
Add rake task 'gem:windows' which makes use of the new rake-
|
270
|
+
compiler-dock.
|
271
|
+
|
272
|
+
Also update the Gemfile per 'rake bundler:gemfile' .
|
273
|
+
[fafadd62cc6d]
|
274
|
+
|
275
|
+
* Rakefile.cross:
|
276
|
+
Cross build: Work around missing dependency to libcommon in
|
277
|
+
PostgreSQL-9.4.0
|
278
|
+
[630a22437026]
|
279
|
+
|
280
|
+
2015-05-25 Michael Granger <ged@FaerieMUD.org>
|
281
|
+
|
282
|
+
* ext/pg_connection.c:
|
283
|
+
Change instance_of checks to kind_of for subclassing [#220]
|
284
|
+
[eb4d3c003bd6] [github/kind-of-fix]
|
285
|
+
|
286
|
+
* lib/pg/basic_type_mapping.rb:
|
287
|
+
Whitespace fixes
|
288
|
+
[4a4368e30bcf] [github/master]
|
289
|
+
|
290
|
+
2015-05-14 Michael Granger <ged@FaerieMUD.org>
|
291
|
+
|
292
|
+
* .hgtags:
|
293
|
+
Added tag v0.18.2 for changeset 22a361201fd1
|
294
|
+
[94e26b472c78]
|
295
|
+
|
296
|
+
* .hgsigs:
|
297
|
+
Added signature for changeset 7d31b04e7913
|
298
|
+
[22a361201fd1] [v0.18.2]
|
299
|
+
|
300
|
+
* History.rdoc, lib/pg.rb:
|
301
|
+
Bump patch version, update History
|
302
|
+
[7d31b04e7913]
|
303
|
+
|
304
|
+
* Merged with git branch.
|
305
|
+
[45a1753da19a]
|
306
|
+
|
307
|
+
* .rvm.gems:
|
308
|
+
Bump the rvm gemset gem versions
|
309
|
+
[8050020330d8]
|
310
|
+
|
311
|
+
2015-03-13 Lars Kanis <lars@greiz-reinsdorf.de>
|
312
|
+
|
313
|
+
* ext/pg_text_encoder.c, ext/util.c, ext/util.h:
|
314
|
+
Avoid possible symbol name clash when linking againt static libpq.
|
315
|
+
|
316
|
+
This should fix issue #209 .
|
317
|
+
[2334a1bc38e6]
|
318
|
+
|
319
|
+
2015-02-11 Lars Kanis <lars@greiz-reinsdorf.de>
|
320
|
+
|
321
|
+
* ext/pg.c, spec/pg/connection_spec.rb:
|
322
|
+
Merge from git branch.
|
323
|
+
[a7261eb6796f]
|
324
|
+
|
325
|
+
* ext/pg.h, ext/pg_text_decoder.c:
|
326
|
+
Merge branch 'non-c99-compiler'
|
327
|
+
[7e8f1e4da59d]
|
328
|
+
|
329
|
+
2015-02-06 Lars Kanis <lars@greiz-reinsdorf.de>
|
330
|
+
|
331
|
+
* ext/extconf.rb, ext/pg.h, ext/pg_result.c, ext/pg_text_decoder.c,
|
332
|
+
ext/pg_type_map_by_mri_type.c:
|
333
|
+
Use fixed size arrays in case the compiler doesn't support C99
|
334
|
+
variable length arrays.
|
335
|
+
|
336
|
+
This is notably the MSVC compiler.
|
337
|
+
|
338
|
+
Note: The MSVC build environment is currently not tested on a
|
339
|
+
regular base, so is not fully supported.
|
340
|
+
[a08290a78221]
|
341
|
+
|
342
|
+
2015-02-11 Lars Kanis <lars@greiz-reinsdorf.de>
|
343
|
+
|
344
|
+
* ext/pg_connection.c:
|
345
|
+
Replace quote_ident implementation with that of
|
346
|
+
PG::TextEncoder::Identifier .
|
347
|
+
|
348
|
+
This allowes quoation of array of strings in addition to plain
|
349
|
+
strings and removes duplicated code.
|
350
|
+
|
351
|
+
It also removes the 63 byte limitation and therefore obsolets pull
|
352
|
+
request #21 https://bitbucket.org/ged/ruby-pg/pull-request/21
|
353
|
+
[b8ac29b020ee]
|
354
|
+
|
355
|
+
* ext/pg.c, ext/pg.h, ext/pg_text_decoder.c, ext/pg_text_encoder.c,
|
356
|
+
spec/pg/connection_spec.rb, spec/pg/type_spec.rb:
|
357
|
+
Derive PG::TextDecoder::Identifier from PG::SimpleDecoder instead of
|
358
|
+
CompositeDecoder
|
359
|
+
|
360
|
+
.. and equally for Encoder.
|
361
|
+
|
362
|
+
Identifiers are strings, so there is no good reason to compose it
|
363
|
+
from other element types. This change allows a faster quoting
|
364
|
+
implementation.
|
365
|
+
[55ec58cabad4]
|
366
|
+
|
367
|
+
2015-05-14 Michael Granger <ged@FaerieMUD.org>
|
368
|
+
|
369
|
+
* .rvmrc:
|
370
|
+
Bump default rvm Ruby to 2.2
|
371
|
+
[128b075d5800]
|
372
|
+
|
373
|
+
2015-03-13 Lars Kanis <lars@greiz-reinsdorf.de>
|
374
|
+
|
375
|
+
* ext/pg_text_encoder.c, ext/util.c, ext/util.h:
|
376
|
+
Avoid possible symbol name clash when linking againt static libpq.
|
377
|
+
|
378
|
+
This should fix issue #209 .
|
379
|
+
[b859963462b2]
|
380
|
+
|
381
|
+
2015-02-11 Lars Kanis <lars@greiz-reinsdorf.de>
|
382
|
+
|
383
|
+
* ext/pg.c, spec/pg/connection_spec.rb:
|
384
|
+
Merge from git branch.
|
385
|
+
[b60c89ee93c8]
|
386
|
+
|
387
|
+
* ext/pg.h, ext/pg_text_decoder.c:
|
388
|
+
Merge branch 'non-c99-compiler'
|
389
|
+
[b111c84b8445]
|
390
|
+
|
391
|
+
2015-02-06 Lars Kanis <lars@greiz-reinsdorf.de>
|
392
|
+
|
393
|
+
* ext/extconf.rb, ext/pg.h, ext/pg_result.c, ext/pg_text_decoder.c,
|
394
|
+
ext/pg_type_map_by_mri_type.c:
|
395
|
+
Use fixed size arrays in case the compiler doesn't support C99
|
396
|
+
variable length arrays.
|
397
|
+
|
398
|
+
This is notably the MSVC compiler.
|
399
|
+
|
400
|
+
Note: The MSVC build environment is currently not tested on a
|
401
|
+
regular base, so is not fully supported.
|
402
|
+
[1269b8ad77b8]
|
403
|
+
|
404
|
+
2015-02-11 Lars Kanis <lars@greiz-reinsdorf.de>
|
405
|
+
|
406
|
+
* ext/pg_connection.c:
|
407
|
+
Replace quote_ident implementation with that of
|
408
|
+
PG::TextEncoder::Identifier .
|
409
|
+
|
410
|
+
This allowes quoation of array of strings in addition to plain
|
411
|
+
strings and removes duplicated code.
|
412
|
+
|
413
|
+
It also removes the 63 byte limitation and therefore obsolets pull
|
414
|
+
request #21 https://bitbucket.org/ged/ruby-pg/pull-request/21
|
415
|
+
[714601d05fba]
|
416
|
+
|
417
|
+
* ext/pg.c, ext/pg.h, ext/pg_text_decoder.c, ext/pg_text_encoder.c,
|
418
|
+
spec/pg/connection_spec.rb, spec/pg/type_spec.rb:
|
419
|
+
Derive PG::TextDecoder::Identifier from PG::SimpleDecoder instead of
|
420
|
+
CompositeDecoder
|
421
|
+
|
422
|
+
.. and equally for Encoder.
|
423
|
+
|
424
|
+
Identifiers are strings, so there is no good reason to compose it
|
425
|
+
from other element types. This change allows a faster quoting
|
426
|
+
implementation.
|
427
|
+
[df67241d1a4f]
|
428
|
+
|
429
|
+
2015-02-03 Michael Granger <ged@FaerieMUD.org>
|
430
|
+
|
431
|
+
* README.rdoc, ext/pg.c:
|
432
|
+
Update copyright notices
|
433
|
+
[447962e0dcba]
|
434
|
+
|
435
|
+
* spec/pg/connection_spec.rb:
|
436
|
+
Whitespace fixes
|
437
|
+
[f4a85f78807d]
|
438
|
+
|
439
|
+
2014-08-22 Chris Bandy <bandy.chris@gmail.com>
|
440
|
+
|
441
|
+
* lib/pg/connection.rb, spec/pg/connection_spec.rb:
|
442
|
+
Allow URI connection string
|
443
|
+
|
444
|
+
(Imported from https://github.com/ged/ruby-pg/pull/3)
|
445
|
+
[d4579ef9eae4]
|
446
|
+
|
447
|
+
2015-01-27 Lars Kanis <lars@greiz-reinsdorf.de>
|
448
|
+
|
449
|
+
* lib/pg/text_decoder.rb, spec/pg/type_spec.rb:
|
450
|
+
Fix decoding of fractional timezones like "-00:25:21".
|
451
|
+
[766aa353b392]
|
452
|
+
|
453
|
+
2015-01-27 Lars Kanis <kanis@comcard.de>
|
454
|
+
|
455
|
+
* History.rdoc, lib/pg/text_decoder.rb, spec/pg/type_spec.rb:
|
456
|
+
Fix decoding of fractional timezones like "04:30".
|
457
|
+
|
458
|
+
This fixes https://bitbucket.org/ged/ruby-pg/issue/203 .
|
459
|
+
[33225805259c]
|
460
|
+
|
461
|
+
2015-01-16 Lars Kanis <kanis@comcard.de>
|
462
|
+
|
463
|
+
* History.rdoc, ext/extconf.rb, ext/pg_connection.c,
|
464
|
+
spec/pg/connection_spec.rb:
|
465
|
+
Revert "Add PG::Connection#hostaddr. This is new in PostgreSQL-9.4."
|
466
|
+
|
467
|
+
PQhostaddr was removed from final PostgreSQL-9.4.0.
|
468
|
+
|
469
|
+
This fixes issue #202 :https://bitbucket.org/ged/ruby-pg/issue/202
|
470
|
+
|
471
|
+
This reverts commit 03deab7aa63f9a00984014ce59617d552b4eac34.
|
472
|
+
[b51d736bca77]
|
473
|
+
|
474
|
+
* History.rdoc:
|
475
|
+
Add a missing enhancement for pg-0.18.0
|
476
|
+
[a7635c3425f3]
|
477
|
+
|
478
|
+
* .travis.yml:
|
479
|
+
Update ruby and postgresql versions for travic-ci
|
480
|
+
[b67b92f7db97]
|
481
|
+
|
482
|
+
2015-01-06 Lars Kanis <kanis@comcard.de>
|
483
|
+
|
484
|
+
* Merge from git.
|
485
|
+
[625e9359ac47]
|
486
|
+
|
487
|
+
2015-01-05 Lars Kanis <kanis@comcard.de>
|
488
|
+
|
489
|
+
* ext/pg_type_map_in_ruby.c:
|
490
|
+
Use a simpler fix for https://bitbucket.org/ged/ruby-
|
491
|
+
pg/commits/a38cf53a96f
|
492
|
+
[3d89d3aae4fd]
|
493
|
+
|
494
|
+
2015-01-05 Michael Granger <ged@FaerieMUD.org>
|
495
|
+
|
496
|
+
* .hgtags:
|
497
|
+
Added tag v0.18.1 for changeset f97dd6cb4f34
|
498
|
+
[f8122014065a]
|
499
|
+
|
500
|
+
* .hgsigs:
|
501
|
+
Added signature for changeset ba5aff64b5cb
|
502
|
+
[f97dd6cb4f34] [v0.18.1]
|
503
|
+
|
504
|
+
* History.rdoc, lib/pg.rb:
|
505
|
+
Bump the patch version, update history
|
506
|
+
[ba5aff64b5cb]
|
507
|
+
|
508
|
+
* Rakefile:
|
509
|
+
Fix the minimum Ruby version (closes #199)
|
510
|
+
[69422cc8dc45]
|
511
|
+
|
512
|
+
2015-01-03 Michael Granger <ged@FaerieMUD.org>
|
513
|
+
|
514
|
+
* .hgtags:
|
515
|
+
Added tag v0.18.0 for changeset cfb2bfc0f661
|
516
|
+
[8aff3973f2d2]
|
517
|
+
|
518
|
+
* .hgsigs:
|
519
|
+
Added signature for changeset c519766e3ec9
|
520
|
+
[cfb2bfc0f661] [v0.18.0]
|
521
|
+
|
522
|
+
* History.rdoc:
|
523
|
+
Update the History
|
524
|
+
[c519766e3ec9]
|
525
|
+
|
526
|
+
* Manifest.txt:
|
527
|
+
Keep the manifest sorted so it can be diffed.
|
528
|
+
[d7cdbfd1b516]
|
529
|
+
|
530
|
+
* Rakefile:
|
531
|
+
Add a Rake task for generating the gemspec
|
532
|
+
[ec6d6ce61a15]
|
533
|
+
|
534
|
+
2014-12-26 Lars Kanis <lars@greiz-reinsdorf.de>
|
535
|
+
|
536
|
+
* lib/pg/basic_type_mapping.rb:
|
537
|
+
Add a warning about Type OIDs without decoder in
|
538
|
+
PG::BasicTypeMapForResults.
|
539
|
+
[592e29cdf2e9]
|
540
|
+
|
541
|
+
2014-12-13 Lars Kanis <lars@greiz-reinsdorf.de>
|
542
|
+
|
543
|
+
* ext/pg_type_map_in_ruby.c:
|
544
|
+
Fix TypeMapInRuby for Rubinius.
|
545
|
+
[a38cf53a96f1]
|
546
|
+
|
547
|
+
2014-12-12 Lars Kanis <lars@greiz-reinsdorf.de>
|
548
|
+
|
549
|
+
* ext/pg_type_map_all_strings.c:
|
550
|
+
Fix documentation to TypeMapAllStrings.
|
551
|
+
[c53f993a4254]
|
552
|
+
|
553
|
+
* Manifest.txt, ext/pg.c, ext/pg.h, ext/pg_type_map_in_ruby.c,
|
554
|
+
spec/pg/type_map_in_ruby_spec.rb:
|
555
|
+
Add TypeMapInRuby as a way to combine fast but static C-based type
|
556
|
+
maps with flexible ruby based type maps.
|
557
|
+
[a9724aef9116]
|
558
|
+
|
559
|
+
* ext/pg_type_map.c, ext/pg_type_map_by_oid.c,
|
560
|
+
lib/pg/basic_type_mapping.rb, spec/pg/basic_type_mapping_spec.rb,
|
561
|
+
spec/pg/type_map_by_oid_spec.rb, spec/pg/type_map_spec.rb:
|
562
|
+
Remove TypeMap#fit_to_result and #fit_to_query.
|
563
|
+
|
564
|
+
The only use case so far is building a TypeMapByColumn out of a
|
565
|
+
TypeMapByOid, which is now named more explicitly as
|
566
|
+
#build_column_map .
|
567
|
+
[c99d26015e3c]
|
568
|
+
|
569
|
+
* ext/pg_text_encoder.c:
|
570
|
+
Remove unused variable.
|
571
|
+
[1a13e7eafeb7]
|
572
|
+
|
573
|
+
* ext/pg_connection.c, ext/pg_copy_coder.c, lib/pg/connection.rb:
|
574
|
+
Several documentation fixes to COPY coders.
|
575
|
+
[a7befacdef4e]
|
576
|
+
|
577
|
+
2014-11-26 Lars Kanis <lars@greiz-reinsdorf.de>
|
578
|
+
|
579
|
+
* ext/pg_text_encoder.c, spec/pg/type_spec.rb:
|
580
|
+
PG::TextDecoder::Boolean: encode only true, false, 0 and 1 and pass
|
581
|
+
all other values as strings to the server.
|
582
|
+
[6b86a5543567]
|
583
|
+
|
584
|
+
2014-11-24 Lars Kanis <kanis@comcard.de>
|
585
|
+
|
586
|
+
* spec/pg/type_map_by_class_spec.rb:
|
587
|
+
Fix test case for rbx. The error message is somewhat different than
|
588
|
+
in MRI.
|
589
|
+
[99638547f8b4]
|
590
|
+
|
591
|
+
2014-11-21 Lars Kanis <lars@greiz-reinsdorf.de>
|
592
|
+
|
593
|
+
* ext/pg_text_decoder.c:
|
594
|
+
Stack memory can be used for Identifier decoding, since they are
|
595
|
+
limited in size.
|
596
|
+
[b7f9daeeba29]
|
597
|
+
|
598
|
+
2014-11-17 Lars Kanis <kanis@comcard.de>
|
599
|
+
|
600
|
+
* .travis.yml:
|
601
|
+
Revert "Remove Rubinius from the allowed failure list - failing
|
602
|
+
specs are now marked as pending."
|
603
|
+
|
604
|
+
This reverts commit 524c84ccd17f3caeed13a16084206f14225bbdd6 because
|
605
|
+
of ongoing issues with gem installation on rbx.
|
606
|
+
[b8debfea4331]
|
607
|
+
|
608
|
+
* ext/pg.h, ext/pg_errors.c, ext/pg_result.c, spec/pg/result_spec.rb:
|
609
|
+
Use dedicated error classes for stream exceptions.
|
610
|
+
[f23dd01bcb52]
|
611
|
+
|
612
|
+
* History.rdoc, ext/pg.c, spec/pg_spec.rb:
|
613
|
+
Add PG.init_openssl and PG.init_ssl methods.
|
614
|
+
|
615
|
+
This will hopefully help to solve issue #190:
|
616
|
+
https://bitbucket.org/ged/ruby-pg/issue/190
|
617
|
+
[45610159a246]
|
618
|
+
|
619
|
+
2014-11-10 Lars Kanis <lars@greiz-reinsdorf.de>
|
620
|
+
|
621
|
+
* ext/pg_type_map_by_class.c:
|
622
|
+
Fix typo.
|
623
|
+
[eeb8a82c5328]
|
624
|
+
|
625
|
+
* spec/pg/type_map_by_class_spec.rb:
|
626
|
+
TypeMapByClass: Add test for invlid coder object.
|
627
|
+
[64ae7cff7f40]
|
628
|
+
|
629
|
+
* ext/pg_type_map_by_class.c:
|
630
|
+
Better check than believe the value type returned by #ancestors .
|
631
|
+
[961a1fce08fd]
|
632
|
+
|
633
|
+
2014-11-08 Lars Kanis <lars@greiz-reinsdorf.de>
|
634
|
+
|
635
|
+
* ext/pg_binary_decoder.c, ext/pg_type_map_all_strings.c:
|
636
|
+
Tweak the type cast documentation a bit more. [skip ci]
|
637
|
+
[185638b52684]
|
638
|
+
|
639
|
+
* ext/pg_type_map_by_class.c, spec/pg/type_map_by_class_spec.rb:
|
640
|
+
TypeMapByClass: Forgot to expire the cache on changes.
|
641
|
+
[1ce48a854d11]
|
642
|
+
|
643
|
+
2014-11-07 Lars Kanis <lars@greiz-reinsdorf.de>
|
644
|
+
|
645
|
+
* Manifest.txt, ext/pg.c, ext/pg.h, ext/pg_type_map_by_class.c,
|
646
|
+
ext/pg_type_map_by_mri_type.c, lib/pg/basic_type_mapping.rb,
|
647
|
+
spec/pg/connection_spec.rb, spec/pg/type_map_by_class_spec.rb,
|
648
|
+
spec/pg/type_map_by_column_spec.rb, spec/pg/type_spec.rb:
|
649
|
+
Add PG::TypeMapByClass. This replaces PG::TypeMapByMriType with a
|
650
|
+
more portable and flexible way to encode values.
|
651
|
+
[27987dbd0b32]
|
652
|
+
|
653
|
+
2014-11-06 Lars Kanis <lars@greiz-reinsdorf.de>
|
654
|
+
|
655
|
+
* ext/pg_connection.c, ext/pg_result.c:
|
656
|
+
Add note about function usage by the sequel_pg gem.
|
657
|
+
[989a6bd6bf0d]
|
658
|
+
|
659
|
+
2014-11-03 Lars Kanis <lars@greiz-reinsdorf.de>
|
660
|
+
|
661
|
+
* ext/pg_result.c:
|
662
|
+
Fix possible segfault, when the GC runs within the loop of
|
663
|
+
pgresult_init_fnames().
|
664
|
+
[017bc722655b]
|
665
|
+
|
666
|
+
2014-11-01 Lars Kanis <lars@greiz-reinsdorf.de>
|
667
|
+
|
668
|
+
* ext/pg.h, ext/pg_coder.c, lib/pg/coder.rb, spec/pg/type_spec.rb:
|
669
|
+
Set default format code for binary coders to 1.
|
670
|
+
[336f694f6362]
|
671
|
+
|
672
|
+
* ext/pg_text_encoder.c, spec/pg/type_spec.rb:
|
673
|
+
Add PG::TextEncoder::Bytea .
|
674
|
+
[b10cbaaead6b]
|
675
|
+
|
676
|
+
2014-10-31 Lars Kanis <lars@greiz-reinsdorf.de>
|
677
|
+
|
678
|
+
* lib/pg.rb:
|
679
|
+
Windows: Set the PATH so that the bundled libpq.dll of the binary
|
680
|
+
gem is loaded preferably to other dlls in the system.
|
681
|
+
|
682
|
+
This avoids loading issues, if there is some other libpq.dll in the
|
683
|
+
path, that fails to load for some reason.
|
684
|
+
|
685
|
+
[skip ci]
|
686
|
+
[b2bf034e3b9f]
|
687
|
+
|
688
|
+
2014-10-30 Lars Kanis <lars@greiz-reinsdorf.de>
|
689
|
+
|
690
|
+
* History.rdoc:
|
691
|
+
Update History.rdoc for PG::Connection#hostaddr .
|
692
|
+
[2d1a83316472]
|
693
|
+
|
694
|
+
* .travis.yml:
|
695
|
+
Remove Rubinius from the allowed failure list - failing specs are
|
696
|
+
now marked as pending.
|
697
|
+
[9e707fcf2070]
|
698
|
+
|
699
|
+
* ext/extconf.rb, ext/pg_connection.c, spec/pg/connection_spec.rb:
|
700
|
+
Add PG::Connection#hostaddr. This is new in PostgreSQL-9.4.
|
701
|
+
[3a7b6a1d1cfe]
|
702
|
+
|
703
|
+
* .travis.yml, spec/helpers.rb:
|
704
|
+
Update tavis-ci and specs for PostgreSQL-9.4.
|
705
|
+
[4daf11be6d3d]
|
706
|
+
|
707
|
+
2014-10-28 Lars Kanis <lars@greiz-reinsdorf.de>
|
708
|
+
|
709
|
+
* spec/pg/type_spec.rb:
|
710
|
+
Fix typo in specs. Thanks Chris Bandy.
|
711
|
+
[9015809803aa]
|
712
|
+
|
713
|
+
2014-10-27 Lars Kanis <lars@greiz-reinsdorf.de>
|
714
|
+
|
715
|
+
* spec/pg/result_spec.rb:
|
716
|
+
Mark failing tests on Rubinius as pending.
|
717
|
+
[12f053154a99]
|
718
|
+
|
719
|
+
* History.rdoc:
|
720
|
+
Update History.rdoc
|
721
|
+
[b7faf8c1c677]
|
722
|
+
|
723
|
+
* Merge branch 'master' of github.com:larskanis/ruby-pg
|
724
|
+
[0f49f9bc93c6]
|
725
|
+
|
726
|
+
2014-10-27 Lars Kanis <kanis@comcard.de>
|
727
|
+
|
728
|
+
* ext/extconf.rb, ext/pg.h:
|
729
|
+
Fix compat with Rubinius.
|
730
|
+
[3a4e19c56686]
|
731
|
+
|
732
|
+
2014-10-27 Lars Kanis <lars@greiz-reinsdorf.de>
|
733
|
+
|
734
|
+
* ext/pg_result.c, spec/pg/result_spec.rb:
|
735
|
+
Add PG::Result#stream_each and #stream_each_row methods and specs.
|
736
|
+
[9387b7aeaad3]
|
737
|
+
|
738
|
+
* ext/pg_result.c:
|
739
|
+
Fix typo.
|
740
|
+
[d2bfa9acf36b]
|
741
|
+
|
742
|
+
2014-10-25 Lars Kanis <lars@greiz-reinsdorf.de>
|
743
|
+
|
744
|
+
* ext/pg.h, ext/pg_result.c:
|
745
|
+
Use rb_hash_dup() with a prepared tuple hash instead of
|
746
|
+
rb_hash_new().
|
747
|
+
|
748
|
+
This gives a little speedup of 5-10% for Result#each .
|
749
|
+
[dfe51787f74d]
|
750
|
+
|
751
|
+
2014-10-24 Lars Kanis <lars@greiz-reinsdorf.de>
|
752
|
+
|
753
|
+
* ext/pg.h:
|
754
|
+
Add workaround for missing RETURN_SIZED_ENUMERATOR macro in Ruby <
|
755
|
+
2.0.
|
756
|
+
[995d0d85d88d]
|
757
|
+
|
758
|
+
* ext/pg_text_encoder.c:
|
759
|
+
Add include <math.h>. It defines isnan() and isinf().
|
760
|
+
[cadf29e3f863]
|
761
|
+
|
762
|
+
* ext/pg_result.c:
|
763
|
+
Use rb_ary_new4() instead of rb_ary_new_from_values() for compat
|
764
|
+
with Ruby < 2.1.
|
765
|
+
[7c465876616d]
|
766
|
+
|
767
|
+
* ext/pg_result.c:
|
768
|
+
Populate the row-Array in Result#each_row and Result.values per
|
769
|
+
rb_ary_new_from_values() from VALUE array on the stack. The number
|
770
|
+
of fields is limited on the server side to ~1700, so that stack
|
771
|
+
space is not critical. This gives a small speed up by 5%.
|
772
|
+
[adc7a83326e3]
|
773
|
+
|
774
|
+
* ext/pg.h, ext/pg_result.c:
|
775
|
+
Use frozen String objects as field names and reuse them where
|
776
|
+
possible.
|
777
|
+
|
778
|
+
This speeds Result#each up by 30%.
|
779
|
+
[f50c049adc3b]
|
780
|
+
|
781
|
+
* ext/pg_result.c, spec/pg/result_spec.rb:
|
782
|
+
Add Enumerator versions to Result#each and #each_row.
|
783
|
+
[1b324be632e4]
|
784
|
+
|
785
|
+
2014-10-22 Lars Kanis <lars@greiz-reinsdorf.de>
|
786
|
+
|
787
|
+
* ext/pg_connection.c:
|
788
|
+
Merge alloc_query_params() and alloc_query_params1(). The separation
|
789
|
+
is no longer needed.
|
790
|
+
|
791
|
+
Allocate gc_array on demand, as it's not needed for most queries.
|
792
|
+
[aae1c7dd2360]
|
793
|
+
|
794
|
+
* README.rdoc, ext/pg.h, ext/pg_connection.c, ext/pg_copy_coder.c,
|
795
|
+
ext/pg_result.c, ext/pg_type_map.c, ext/pg_type_map_all_strings.c,
|
796
|
+
ext/pg_type_map_by_column.c, ext/pg_type_map_by_mri_type.c,
|
797
|
+
ext/pg_type_map_by_oid.c, spec/pg/basic_type_mapping_spec.rb,
|
798
|
+
spec/pg/connection_spec.rb, spec/pg/result_spec.rb,
|
799
|
+
spec/pg/type_map_by_column_spec.rb,
|
800
|
+
spec/pg/type_map_by_mri_type_spec.rb,
|
801
|
+
spec/pg/type_map_by_oid_spec.rb:
|
802
|
+
Allow cascading of type maps.
|
803
|
+
|
804
|
+
If a value can not be handled by a type map this value isn't handled
|
805
|
+
implicit by TypeMapAllStrings, but instead by the assigned
|
806
|
+
default_type_map, now.
|
807
|
+
[d369d31e8fe3]
|
808
|
+
|
809
|
+
2014-10-17 Lars Kanis <lars@greiz-reinsdorf.de>
|
810
|
+
|
811
|
+
* ext/pg.h, ext/pg_type_map.c, ext/pg_type_map_by_column.c,
|
812
|
+
spec/pg/type_map_by_column_spec.rb:
|
813
|
+
Gracefully handle not initialized state for PG::TypeMapByColumn
|
814
|
+
derivations.
|
815
|
+
[485d02650e09]
|
816
|
+
|
817
|
+
2014-10-15 Lars Kanis <lars@greiz-reinsdorf.de>
|
818
|
+
|
819
|
+
* ext/pg_text_encoder.c, spec/pg/type_spec.rb:
|
820
|
+
Allow non Array values as input for TextEncoder::Array.
|
821
|
+
[a4725dfca9e4]
|
822
|
+
|
823
|
+
* ext/pg_coder.c, spec/pg/type_spec.rb:
|
824
|
+
Clarify handling of nil/NULL values by PG::Coders.
|
825
|
+
[99d6acb8b66d]
|
826
|
+
|
827
|
+
* ext/pg_text_encoder.c, ext/util.c, ext/util.h:
|
828
|
+
Move pg_strncasecmp() to util.c - it better fits there.
|
829
|
+
[117fb5c5eed7]
|
830
|
+
|
831
|
+
2014-10-14 Lars Kanis <lars@greiz-reinsdorf.de>
|
832
|
+
|
833
|
+
* ext/pg_text_encoder.c, spec/pg/type_spec.rb:
|
834
|
+
Implement a more serious TestEncoder::Boolean.
|
835
|
+
[9063a84a2ff9]
|
836
|
+
|
837
|
+
* ext/pg.h, ext/pg_binary_encoder.c, ext/pg_coder.c,
|
838
|
+
ext/pg_text_encoder.c:
|
839
|
+
Fix naming and description of pg_coder_enc_to_s(). It actually uses
|
840
|
+
#to_s not #to_str.
|
841
|
+
[ac23631c96d9]
|
842
|
+
|
843
|
+
* ext/pg_text_encoder.c, spec/pg/type_spec.rb:
|
844
|
+
Use same rules for array quoting in text encoder as PostgreSQL.
|
845
|
+
|
846
|
+
This makes mutate-detection easier (in Rails).
|
847
|
+
[9be31c0663e3]
|
848
|
+
|
849
|
+
2014-10-12 Lars Kanis <lars@greiz-reinsdorf.de>
|
850
|
+
|
851
|
+
* ext/pg_binary_decoder.c:
|
852
|
+
Remove needless memcpy() call.
|
853
|
+
[f786006b25ff]
|
854
|
+
|
855
|
+
* ext/pg_binary_decoder.c, ext/pg_binary_encoder.c, ext/util.c,
|
856
|
+
ext/util.h:
|
857
|
+
Change the endianess conversion for binary data from a library issue
|
858
|
+
to a compiler issue.
|
859
|
+
|
860
|
+
This removes dependency to beXXtoh() functions and ensures aligned
|
861
|
+
memory access.
|
862
|
+
[c8d7c26dd595]
|
863
|
+
|
864
|
+
2014-10-11 Michael Granger <ged@FaerieMUD.org>
|
865
|
+
|
866
|
+
* ext/extconf.rb, spec/helpers.rb, spec/pg/connection_spec.rb:
|
867
|
+
Merge pull request #6 from larskanis/column_type_mapping
|
868
|
+
|
869
|
+
Add type mapping and performance improvements
|
870
|
+
[0da90c2b2812]
|
871
|
+
|
872
|
+
2014-10-09 Lars Kanis <lars@greiz-reinsdorf.de>
|
873
|
+
|
874
|
+
* Rakefile, Rakefile.cross, ext/extconf.rb:
|
875
|
+
gcc option -static-libgcc is required for i386-mingw32 build.
|
876
|
+
|
877
|
+
Don't use native pg_config for cross build. The paths were
|
878
|
+
overwritten anyways until now.
|
879
|
+
|
880
|
+
Remove code that was only needed for static build to libpq.
|
881
|
+
[9a2f9771d0b6]
|
882
|
+
|
883
|
+
2014-10-10 Lars Kanis <lars@greiz-reinsdorf.de>
|
884
|
+
|
885
|
+
* ext/pg_type_map_by_column.c:
|
886
|
+
Use C99 struct initializer syntax.
|
887
|
+
[ec9ef7fb39eb]
|
888
|
+
|
889
|
+
2014-10-09 Lars Kanis <lars@greiz-reinsdorf.de>
|
890
|
+
|
891
|
+
* spec/pg/connection_spec.rb, spec/pg/type_spec.rb:
|
892
|
+
Fix specs for Windows.
|
893
|
+
[3659cfd75b96]
|
894
|
+
|
895
|
+
* .travis.yml, ext/pg_connection.c, lib/pg.rb,
|
896
|
+
spec/pg/connection_spec.rb:
|
897
|
+
Merge branch 'master' into column_type_mapping
|
898
|
+
|
899
|
+
Conflicts: ext/pg_connection.c spec/pg/connection_spec.rb
|
900
|
+
[57d770944b5d]
|
901
|
+
|
902
|
+
* ext/pg.h, ext/pg_copy_coder.c, ext/pg_text_encoder.c:
|
903
|
+
Fix gcc warnings in mingw build.
|
904
|
+
[26541f7e6bbb]
|
905
|
+
|
906
|
+
2014-10-07 Lars Kanis <lars@greiz-reinsdorf.de>
|
907
|
+
|
908
|
+
* ext/pg_type_map.c, ext/pg_type_map_all_strings.c,
|
909
|
+
ext/pg_type_map_by_column.c, ext/pg_type_map_by_mri_type.c,
|
910
|
+
ext/pg_type_map_by_oid.c, lib/pg/type_map_by_column.rb:
|
911
|
+
Add documentation to TypeMap classes.
|
912
|
+
[4227fdc5f0ac]
|
913
|
+
|
914
|
+
2014-10-06 Lars Kanis <kanis@comcard.de>
|
915
|
+
|
916
|
+
* ext/pg_connection.c, ext/pg_result.c, ext/pg_type_map_by_mri_type.c:
|
917
|
+
Ensure zero terminated strings per StringValueCStr(), where C-str
|
918
|
+
pointers are expected.
|
919
|
+
|
920
|
+
This fixes issues with ruby-2.2.0-dev described in github pull
|
921
|
+
request #5: https://github.com/ged/ruby-pg/pull/5
|
922
|
+
[f0b7f99b1dd5]
|
923
|
+
|
924
|
+
2014-10-05 Lars Kanis <lars@greiz-reinsdorf.de>
|
925
|
+
|
926
|
+
* ext/pg_binary_decoder.c, ext/pg_binary_encoder.c, ext/pg_coder.c,
|
927
|
+
ext/pg_copy_coder.c, ext/pg_text_decoder.c, ext/pg_text_encoder.c,
|
928
|
+
lib/pg/coder.rb, lib/pg/text_encoder.rb:
|
929
|
+
Add class and method documentation related to type casting.
|
930
|
+
|
931
|
+
Move previous documentaion from .rb files to .c.
|
932
|
+
[e64618a44912]
|
933
|
+
|
934
|
+
* ext/pg_text_encoder.c, spec/pg/type_spec.rb:
|
935
|
+
Format infinity and nan values equally to Float#to_s. This fixes a
|
936
|
+
test case in activerecord.
|
937
|
+
[02285c675326]
|
938
|
+
|
939
|
+
* lib/pg/text_decoder.rb, lib/pg/text_encoder.rb,
|
940
|
+
spec/pg/basic_type_mapping_spec.rb:
|
941
|
+
Support infinity time and date values. String values which can not
|
942
|
+
be parsed/formatted are now passed through instead of raising an
|
943
|
+
exception.
|
944
|
+
[85ce8d239d23]
|
945
|
+
|
946
|
+
2014-10-04 Lars Kanis <lars@greiz-reinsdorf.de>
|
947
|
+
|
948
|
+
* ext/pg_text_decoder.c, ext/pg_text_encoder.c:
|
949
|
+
Add some more documentaion to the encoding/decoding function
|
950
|
+
interface.
|
951
|
+
[47ccb70c45d6]
|
952
|
+
|
953
|
+
* ext/pg.h, ext/pg_coder.c, ext/pg_connection.c, ext/pg_copy_coder.c,
|
954
|
+
ext/pg_text_encoder.c, ext/pg_type_map.c,
|
955
|
+
ext/pg_type_map_all_strings.c, ext/pg_type_map_by_column.c,
|
956
|
+
ext/pg_type_map_by_mri_type.c, spec/pg/type_map_by_column_spec.rb:
|
957
|
+
Fix handling of query parameters in hash form in conjunction with
|
958
|
+
type maps.
|
959
|
+
|
960
|
+
This moves the oid/format determining back into
|
961
|
+
alloc_query_params1() and avoids code duplication within type maps.
|
962
|
+
[2afa6cc41d55]
|
963
|
+
|
964
|
+
* spec/pg/connection_spec.rb:
|
965
|
+
Add missing spec for hash form parameters to #exec_params .
|
966
|
+
[23ad5d676b53]
|
967
|
+
|
968
|
+
2014-10-03 Lars Kanis <lars@greiz-reinsdorf.de>
|
969
|
+
|
970
|
+
* ext/pg_binary_decoder.c, ext/pg_binary_encoder.c,
|
971
|
+
ext/pg_text_decoder.c, ext/pg_text_encoder.c, spec/pg/type_spec.rb:
|
972
|
+
FromBase64 encoder better fits into PG::BinaryEncoder and ToBase64
|
973
|
+
decoder better into PG::BinaryDecoder.
|
974
|
+
[5db80210a077]
|
975
|
+
|
976
|
+
2014-10-02 Lars Kanis <kanis@comcard.de>
|
977
|
+
|
978
|
+
* ext/pg_coder.c, ext/pg_connection.c, ext/pg_text_decoder.c,
|
979
|
+
ext/pg_text_encoder.c:
|
980
|
+
rb_str_new() with NULL ptr can be used to allocate/set the Sting
|
981
|
+
size.
|
982
|
+
[2c4faf369dce]
|
983
|
+
|
984
|
+
* ext/util.c:
|
985
|
+
Speed up the base 64 decoder by 80%.
|
986
|
+
[adb1728e29d9]
|
987
|
+
|
988
|
+
2014-10-01 Lars Kanis <lars@greiz-reinsdorf.de>
|
989
|
+
|
990
|
+
* ext/pg_text_decoder.c:
|
991
|
+
Use strings in base64 decoders, so that we can avoid copy'ing data
|
992
|
+
at string creation in dec_func.
|
993
|
+
[1e2cd139e6fe]
|
994
|
+
|
995
|
+
* spec/pg/type_spec.rb:
|
996
|
+
Fix base64 specs for Ruby < 2.1.
|
997
|
+
[3a9e0cea7387]
|
998
|
+
|
999
|
+
2014-10-01 Lars Kanis <kanis@comcard.de>
|
1000
|
+
|
1001
|
+
* Merge branch 'column_type_mapping' of https://github.com/larskanis
|
1002
|
+
/ruby-pg into column_type_mapping
|
1003
|
+
[712224085b59]
|
1004
|
+
|
1005
|
+
2014-09-30 Lars Kanis <lars@greiz-reinsdorf.de>
|
1006
|
+
|
1007
|
+
* ext/pg.h, ext/pg_result.c, ext/pg_type_map.c,
|
1008
|
+
ext/pg_type_map_all_strings.c, ext/pg_type_map_by_column.c,
|
1009
|
+
ext/pg_type_map_by_oid.c:
|
1010
|
+
Store a pointer to the current typemap into result data and use it
|
1011
|
+
through the this pointer at all places.
|
1012
|
+
|
1013
|
+
This gives PG::Result#getvalue a speedup around 10%.
|
1014
|
+
[a8b3ef723c22]
|
1015
|
+
|
1016
|
+
* spec/pg/basic_type_mapping_spec.rb:
|
1017
|
+
Change timestamp specs to a value before 1970, so that we see, there
|
1018
|
+
is no epoch limitation.
|
1019
|
+
[441da0a4129a]
|
1020
|
+
|
1021
|
+
2014-09-30 Lars Kanis <kanis@comcard.de>
|
1022
|
+
|
1023
|
+
* ext/pg_text_decoder.c, ext/pg_text_encoder.c, ext/util.c,
|
1024
|
+
ext/util.h, spec/pg/type_spec.rb:
|
1025
|
+
Add ToBase64 and FromBase64 classes as encoders and decoders.
|
1026
|
+
[9ff6f35497ad]
|
1027
|
+
|
1028
|
+
2014-09-29 Lars Kanis <kanis@comcard.de>
|
1029
|
+
|
1030
|
+
* ext/pg_connection.c:
|
1031
|
+
Add encoder and decoder for copy_data to gc mark list.
|
1032
|
+
[a445b16a259e]
|
1033
|
+
|
1034
|
+
* Manifest.txt:
|
1035
|
+
Update Manifest.txt
|
1036
|
+
[e13ff4f6433e]
|
1037
|
+
|
1038
|
+
* ext/pg_connection.c:
|
1039
|
+
Add new function pg_get_connection_safe() and use it instead of two
|
1040
|
+
calls to pg_get_connection() and pg_get_pgconn(). It is faster and
|
1041
|
+
less confusing.
|
1042
|
+
[6ad4490a691b]
|
1043
|
+
|
1044
|
+
* ext/pg_coder.c, ext/pg_copy_coder.c:
|
1045
|
+
Fix tainted mark for PG::Coder#encoder and #decode and
|
1046
|
+
PG::TextDecoder::CopyRow output.
|
1047
|
+
[7aec08b7677c]
|
1048
|
+
|
1049
|
+
2014-09-27 Lars Kanis <lars@greiz-reinsdorf.de>
|
1050
|
+
|
1051
|
+
* lib/pg/basic_type_mapping.rb:
|
1052
|
+
Fix PG::BasicTypeRegistry, so that it works with and without
|
1053
|
+
previously applied type mapping.
|
1054
|
+
[db5c2973f143]
|
1055
|
+
|
1056
|
+
* lib/pg/basic_type_mapping.rb:
|
1057
|
+
Underscore unused variables for now, to avoid warnings about them.
|
1058
|
+
[8d016bc21b9a]
|
1059
|
+
|
1060
|
+
* README.rdoc:
|
1061
|
+
Mention type maps for COPY in the README.
|
1062
|
+
[ac361f811253]
|
1063
|
+
|
1064
|
+
* ext/pg.h:
|
1065
|
+
It's better to check for availability of the INLINED macro than for
|
1066
|
+
RUBINIUS define.
|
1067
|
+
[9ba559feacac]
|
1068
|
+
|
1069
|
+
* lib/pg/basic_type_mapping.rb, spec/pg/basic_type_mapping_spec.rb:
|
1070
|
+
Add PG::BasicTypeMapBasedOnResult and add specs for deriving
|
1071
|
+
TypeMapByColumn from a result and use it with COPY.
|
1072
|
+
[6c69768961c1]
|
1073
|
+
|
1074
|
+
* ext/pg_connection.c, ext/pg_result.c:
|
1075
|
+
Add somewhat more documentaion to type casting.
|
1076
|
+
|
1077
|
+
Harmonize coder setters for COPY to accept any PG::Coder (even if
|
1078
|
+
any other than PG::CopyCoder hardly makes sense).
|
1079
|
+
[e0a8ec6d2a04]
|
1080
|
+
|
1081
|
+
* spec/pg/connection_spec.rb:
|
1082
|
+
Mark failing test on Rubinius as pending.
|
1083
|
+
[6eb0fcab4d8c]
|
1084
|
+
|
1085
|
+
* ext/pg_connection.c, spec/pg/connection_spec.rb:
|
1086
|
+
Add typecasting to Connection#get_copy_data.
|
1087
|
+
[21451bfb3659]
|
1088
|
+
|
1089
|
+
* ext/pg.c, ext/pg.h, ext/pg_copy_coder.c, ext/pg_text_encoder.c:
|
1090
|
+
Allow inline checks for string capacity. This removes the bottleneck
|
1091
|
+
in the COPY decoder, when doing the check for each character.
|
1092
|
+
[c2e817ef9b8c]
|
1093
|
+
|
1094
|
+
* ext/pg.c, ext/pg.h, ext/pg_coder.c, ext/pg_connection.c,
|
1095
|
+
ext/pg_copy_coder.c, ext/pg_result.c, ext/pg_text_encoder.c,
|
1096
|
+
ext/pg_type_map.c, ext/pg_type_map_all_strings.c,
|
1097
|
+
ext/pg_type_map_by_column.c, ext/pg_type_map_by_mri_type.c,
|
1098
|
+
spec/pg/type_spec.rb:
|
1099
|
+
Add PG::TypeMapAllStrings and use it as default typemap, when no
|
1100
|
+
explicit typemap was set. This allowes PG::CopyCoder to work without
|
1101
|
+
typemap, too.
|
1102
|
+
[b183ef46969c]
|
1103
|
+
|
1104
|
+
* ext/pg_copy_coder.c, lib/pg/coder.rb, spec/pg/type_spec.rb:
|
1105
|
+
Add CopyCoder#null_string to gc_mark and #to_h .
|
1106
|
+
[893009c81aec]
|
1107
|
+
|
1108
|
+
* ext/pg.h, ext/pg_copy_coder.c, ext/pg_type_map.c,
|
1109
|
+
ext/pg_type_map_by_column.c, ext/pg_type_map_by_mri_type.c,
|
1110
|
+
ext/pg_type_map_by_oid.c, spec/pg/type_spec.rb:
|
1111
|
+
Add decoder for data received from get_copy_data.
|
1112
|
+
|
1113
|
+
Remove duplicated not-suitable-raise-functions.
|
1114
|
+
[0c7c819d2f46]
|
1115
|
+
|
1116
|
+
* ext/pg_copy_coder.c:
|
1117
|
+
Add PG::CopyCoder#null_string .
|
1118
|
+
[3f52c343f9c8]
|
1119
|
+
|
1120
|
+
* ext/pg.c:
|
1121
|
+
Fix description.
|
1122
|
+
[38c8b9d03bd1]
|
1123
|
+
|
1124
|
+
2014-09-26 Lars Kanis <kanis@comcard.de>
|
1125
|
+
|
1126
|
+
* ext/pg.h, ext/pg_coder.c, ext/pg_text_decoder.c,
|
1127
|
+
ext/pg_text_encoder.c, ext/pg_type_map_by_column.c,
|
1128
|
+
ext/pg_type_map_by_oid.c:
|
1129
|
+
Refactured code to find the right decoder function.
|
1130
|
+
[5822d67d8c3b]
|
1131
|
+
|
1132
|
+
* spec/pg/type_spec.rb:
|
1133
|
+
Add decoder specs for Bytea.
|
1134
|
+
[b815372dd5d2]
|
1135
|
+
|
1136
|
+
* lib/pg/coder.rb, spec/pg/type_spec.rb:
|
1137
|
+
Add missing PG::CopyCoder#to_h .
|
1138
|
+
[e0bf32c9eb31]
|
1139
|
+
|
1140
|
+
* ext/pg_copy_coder.c:
|
1141
|
+
Add missing GC-mark function to copycoder.
|
1142
|
+
[70ae4e58761c]
|
1143
|
+
|
1144
|
+
2014-09-25 Lars Kanis <lars@greiz-reinsdorf.de>
|
1145
|
+
|
1146
|
+
* ext/pg_connection.c, lib/pg/connection.rb,
|
1147
|
+
spec/pg/connection_spec.rb:
|
1148
|
+
Implement Connection#put_copy_data and #copy_data with encoder.
|
1149
|
+
[7bfcfe35344c]
|
1150
|
+
|
1151
|
+
* spec/pg/type_spec.rb:
|
1152
|
+
Add more specs for PG::TextEncoder::CopyRow.
|
1153
|
+
[94d6505a6a49]
|
1154
|
+
|
1155
|
+
* ext/pg_coder.c, ext/pg_text_decoder.c, ext/pg_text_encoder.c:
|
1156
|
+
Rename different namings of the instance pointer to "this".
|
1157
|
+
[c07d02ae1cb4]
|
1158
|
+
|
1159
|
+
2014-09-25 Lars Kanis <kanis@comcard.de>
|
1160
|
+
|
1161
|
+
* ext/pg_connection.c:
|
1162
|
+
Save a call to rb_protect by wrapping possibly allocated heap memory
|
1163
|
+
into T_DATA objects that are free'd by the GC.
|
1164
|
+
[3c1d86fedf1b]
|
1165
|
+
|
1166
|
+
* spec/pg/connection_spec.rb:
|
1167
|
+
Mark the failing COPY tests as pending.
|
1168
|
+
[722d4e35b570]
|
1169
|
+
|
1170
|
+
* ext/pg.h, ext/pg_coder.c, ext/pg_copy_coder.c,
|
1171
|
+
ext/pg_text_encoder.c:
|
1172
|
+
Remove duplicated code.
|
1173
|
+
[bf96b3d34538]
|
1174
|
+
|
1175
|
+
* ext/pg_coder.c:
|
1176
|
+
Remove another usage of rb_str_modify_expand.
|
1177
|
+
[36493e777910]
|
1178
|
+
|
1179
|
+
* ext/extconf.rb, ext/pg.c:
|
1180
|
+
Add a more portable version of pg_ensure_str_capa, that should work
|
1181
|
+
on Rubinius.
|
1182
|
+
[bfdd98d3553a]
|
1183
|
+
|
1184
|
+
* ext/pg.c, ext/pg.h, ext/pg_copy_coder.c, ext/pg_text_encoder.c,
|
1185
|
+
spec/pg/type_spec.rb:
|
1186
|
+
Add specs for PG::TextEncoder::CopyRow. Refacture
|
1187
|
+
pg_ensure_str_capa() to avoid duplicated code.
|
1188
|
+
[58db6bcc673e]
|
1189
|
+
|
1190
|
+
2014-09-24 Lars Kanis <lars@greiz-reinsdorf.de>
|
1191
|
+
|
1192
|
+
* ext/pg_text_encoder.c:
|
1193
|
+
Finish second option for encoders to return a string.
|
1194
|
+
[d523e7482832]
|
1195
|
+
|
1196
|
+
2014-09-24 Lars Kanis <kanis@comcard.de>
|
1197
|
+
|
1198
|
+
* ext/pg_coder.c, ext/pg_connection.c, ext/pg_copy_coder.c,
|
1199
|
+
ext/pg_text_encoder.c:
|
1200
|
+
Add a second option for encoders to return a string. Not yet
|
1201
|
+
finished.
|
1202
|
+
[6a408664b53d]
|
1203
|
+
|
1204
|
+
* ext/pg_connection.c:
|
1205
|
+
Fix typo.
|
1206
|
+
[394b095b06e7]
|
1207
|
+
|
1208
|
+
2014-09-23 Lars Kanis <lars@greiz-reinsdorf.de>
|
1209
|
+
|
1210
|
+
* Merge branch 'column_type_mapping' of github.com:larskanis/ruby-pg
|
1211
|
+
into column_type_mapping
|
1212
|
+
[0a904a023886]
|
1213
|
+
|
1214
|
+
2014-09-23 Lars Kanis <kanis@comcard.de>
|
1215
|
+
|
1216
|
+
* ext/pg.h:
|
1217
|
+
Merge branch 'column_type_mapping' of https://github.com/larskanis
|
1218
|
+
/ruby-pg into column_type_mapping
|
1219
|
+
[5f7ae7085714]
|
1220
|
+
|
1221
|
+
* ext/pg_connection.c:
|
1222
|
+
Fix comment.
|
1223
|
+
[bc848e8d054b]
|
1224
|
+
|
1225
|
+
* ext/pg.h, ext/pg_connection.c:
|
1226
|
+
Add setter and getter for Connection#encoder_for_put_copy_data and
|
1227
|
+
decoder_for_get_copy_data.
|
1228
|
+
[b21c2b78af51]
|
1229
|
+
|
1230
|
+
* ext/pg.c, ext/pg.h, ext/pg_binary_encoder.c, ext/pg_copy_coder.c,
|
1231
|
+
ext/pg_text_decoder.c, ext/pg_text_encoder.c,
|
1232
|
+
spec/pg/connection_spec.rb:
|
1233
|
+
Add TextEncoder::CopyRow. Not finished yet.
|
1234
|
+
[f631bec1aea1]
|
1235
|
+
|
1236
|
+
* ext/pg_coder.c:
|
1237
|
+
Refactured duplicated init code for coders.
|
1238
|
+
[89c888335657]
|
1239
|
+
|
1240
|
+
2014-09-23 Lars Kanis <lars@greiz-reinsdorf.de>
|
1241
|
+
|
1242
|
+
* ext/pg_type_map_by_column.c, ext/pg_type_map_by_oid.c:
|
1243
|
+
Fix possible invalid memory access in pg_tmbc_mark().
|
1244
|
+
[89a8049d3de8]
|
1245
|
+
|
1246
|
+
2014-09-22 Lars Kanis <lars@greiz-reinsdorf.de>
|
1247
|
+
|
1248
|
+
* ext/pg.h:
|
1249
|
+
Add compat for Rubinius.
|
1250
|
+
[1836d0c5c623]
|
1251
|
+
|
1252
|
+
2014-09-22 Lars Kanis <kanis@comcard.de>
|
1253
|
+
|
1254
|
+
* ext/pg.h, ext/pg_connection.c, ext/pg_result.c, ext/pg_type_map.c,
|
1255
|
+
ext/pg_type_map_by_column.c, ext/pg_type_map_by_mri_type.c,
|
1256
|
+
ext/pg_type_map_by_oid.c, spec/pg/type_map_by_oid_spec.rb,
|
1257
|
+
spec/pg/type_map_spec.rb:
|
1258
|
+
Add PG::TypeMap#fit_to_result and fit_to_query as Ruby methods and
|
1259
|
+
remove opposite Ruby callback wrapper.
|
1260
|
+
|
1261
|
+
The callback wrapper were not used and explicit calling
|
1262
|
+
fit_to_result makes more sense (for caching etc.).
|
1263
|
+
|
1264
|
+
Swap parameter order of fit_to_result and fit_to_query so that self
|
1265
|
+
is the first one.
|
1266
|
+
[531ece2a12f8]
|
1267
|
+
|
1268
|
+
* ext/pg_type_map_by_mri_type.c:
|
1269
|
+
Add comment to TypeMapByMriType class. [skip-ci]
|
1270
|
+
[19e5622fd318]
|
1271
|
+
|
1272
|
+
* ext/pg_connection.c:
|
1273
|
+
Put only converted ruby strings into the gc_array.
|
1274
|
+
[f705548a5b5f]
|
1275
|
+
|
1276
|
+
2014-09-21 Lars Kanis <lars@greiz-reinsdorf.de>
|
1277
|
+
|
1278
|
+
* README.rdoc, lib/pg/type_map_by_column.rb:
|
1279
|
+
Update the introduction for type casting in Pg. Luckily it became
|
1280
|
+
easier.
|
1281
|
+
[b1876686a1e2]
|
1282
|
+
|
1283
|
+
* lib/pg/basic_type_mapping.rb, spec/pg/basic_type_mapping_spec.rb,
|
1284
|
+
spec/pg/connection_spec.rb:
|
1285
|
+
Use text format for integers and don't set type OID for text types.
|
1286
|
+
[34b9dacb04d3]
|
1287
|
+
|
1288
|
+
* ext/pg_text_encoder.c, spec/pg/type_spec.rb:
|
1289
|
+
Implement a faster Fixnum to text encoder.
|
1290
|
+
[a7b9ea6dffd7]
|
1291
|
+
|
1292
|
+
* spec/pg/connection_spec.rb:
|
1293
|
+
Add spec for a lot of result columns in conjunction with type
|
1294
|
+
casting.
|
1295
|
+
[13629c7af842]
|
1296
|
+
|
1297
|
+
* ext/pg_connection.c:
|
1298
|
+
Use query stack memory pool somewhat more space efficient.
|
1299
|
+
[13c7d27cf253]
|
1300
|
+
|
1301
|
+
2014-09-20 Lars Kanis <lars@greiz-reinsdorf.de>
|
1302
|
+
|
1303
|
+
* ext/pg_connection.c:
|
1304
|
+
Do two passes of query param encoding one after another. This should
|
1305
|
+
give somewhat more cache locality.
|
1306
|
+
[c2634b894cb7]
|
1307
|
+
|
1308
|
+
* ext/pg.h, ext/pg_binary_decoder.c, ext/pg_connection.c,
|
1309
|
+
ext/pg_result.c, ext/pg_text_decoder.c:
|
1310
|
+
Use a faster version of ENCODING_SET.
|
1311
|
+
|
1312
|
+
Remove several ifdefs.
|
1313
|
+
[90051cb74b20]
|
1314
|
+
|
1315
|
+
* ext/pg.c, ext/pg.h, ext/pg_binary_decoder.c, ext/pg_connection.c,
|
1316
|
+
ext/pg_result.c:
|
1317
|
+
Cleanup and harmonize encoding assignment.
|
1318
|
+
[2549c08e631e]
|
1319
|
+
|
1320
|
+
* ext/pg_result.c, spec/pg/result_spec.rb:
|
1321
|
+
Add type casting to Result#column_values and #field_values.
|
1322
|
+
[34b9016757c4]
|
1323
|
+
|
1324
|
+
* ext/pg.h, ext/pg_connection.c, ext/pg_result.c,
|
1325
|
+
ext/pg_text_decoder.c, ext/pg_type_map.c,
|
1326
|
+
ext/pg_type_map_by_column.c, ext/pg_type_map_by_mri_type.c,
|
1327
|
+
ext/pg_type_map_by_oid.c:
|
1328
|
+
Use RBasic encoding store as cache for connection's
|
1329
|
+
internal_encoding.
|
1330
|
+
|
1331
|
+
Remove encoding_index from typemap struct.
|
1332
|
+
|
1333
|
+
Rename typemap-struct member 'typecast' to 'typecast_result_value'.
|
1334
|
+
[58dff8991613]
|
1335
|
+
|
1336
|
+
2014-09-19 Lars Kanis <lars@greiz-reinsdorf.de>
|
1337
|
+
|
1338
|
+
* ext/pg_connection.c:
|
1339
|
+
Avoid duplicated type check for typemaps on query params.
|
1340
|
+
[e7ae19b8fe19]
|
1341
|
+
|
1342
|
+
* ext/pg_type_map_by_mri_type.c:
|
1343
|
+
Sort MRI-types according to their probability.
|
1344
|
+
[383da4885164]
|
1345
|
+
|
1346
|
+
* ext/pg_connection.c, ext/pg_result.c, lib/pg/result.rb:
|
1347
|
+
Implement PG::Result#values in C. This avoids the overhead of
|
1348
|
+
Enumerator-fibers and dynamic array size changes.
|
1349
|
+
[7bf8d41cbcb6]
|
1350
|
+
|
1351
|
+
* ext/pg_connection.c, spec/pg/connection_spec.rb:
|
1352
|
+
Use stack memory for queries unless there are a lot of params.
|
1353
|
+
[b9f9f33f5c21]
|
1354
|
+
|
1355
|
+
2014-09-19 Lars Kanis <kanis@comcard.de>
|
1356
|
+
|
1357
|
+
* ext/pg.h, ext/pg_connection.c, ext/pg_result.c,
|
1358
|
+
ext/pg_type_map_by_column.c, ext/pg_type_map_by_oid.c,
|
1359
|
+
spec/pg/connection_spec.rb, spec/pg/result_spec.rb:
|
1360
|
+
Use a wrapped C-struct as data of PG::Result and store instance
|
1361
|
+
variables there.
|
1362
|
+
|
1363
|
+
This avoids all hash lookups for accessing instance variables.
|
1364
|
+
|
1365
|
+
This also fixes bitbucket bug #185 by marking the result as
|
1366
|
+
autocleared: https://bitbucket.org/ged/ruby-pg/issue/185
|
1367
|
+
|
1368
|
+
And this adds two new methods to PG::Result : autoclear? and
|
1369
|
+
cleared?
|
1370
|
+
[93d51e75cb0b]
|
1371
|
+
|
1372
|
+
* ext/pg_connection.c:
|
1373
|
+
Rename p_conn to this.
|
1374
|
+
[7ae20e3f9db2]
|
1375
|
+
|
1376
|
+
* ext/pg.h, ext/pg_connection.c, ext/pg_result.c:
|
1377
|
+
Move all ivars of the Connection object to C-struct.
|
1378
|
+
[79ed6d2ba9a1]
|
1379
|
+
|
1380
|
+
* ext/pg.h, ext/pg_connection.c:
|
1381
|
+
Merge branch 'column_type_mapping' into use_structs_as_data_ptr
|
1382
|
+
[183eff440b6e]
|
1383
|
+
|
1384
|
+
* ext/pg_connection.c:
|
1385
|
+
Merge branch 'column_type_mapping' of https://github.com/larskanis
|
1386
|
+
/ruby-pg into column_type_mapping
|
1387
|
+
[75ef0371c4ed]
|
1388
|
+
|
1389
|
+
2014-09-18 Lars Kanis <lars@greiz-reinsdorf.de>
|
1390
|
+
|
1391
|
+
* ext/pg_type_map_by_oid.c:
|
1392
|
+
Merge branch 'column_type_mapping' of github.com:larskanis/ruby-pg
|
1393
|
+
into column_type_mapping
|
1394
|
+
|
1395
|
+
Conflicts: ext/pg_type_map_by_oid.c
|
1396
|
+
[39b58e42efd6]
|
1397
|
+
|
1398
|
+
* ext/pg.h, ext/pg_connection.c, ext/pg_type_map.c,
|
1399
|
+
ext/pg_type_map_by_column.c, ext/pg_type_map_by_mri_type.c,
|
1400
|
+
ext/pg_type_map_by_oid.c:
|
1401
|
+
Refacture type casting of query params, to avoid doublicated code.
|
1402
|
+
|
1403
|
+
This also bundles memory allocation for query functions.
|
1404
|
+
[4d0a26946895]
|
1405
|
+
|
1406
|
+
2014-09-18 Lars Kanis <kanis@comcard.de>
|
1407
|
+
|
1408
|
+
* ext/pg.h, ext/pg_connection.c:
|
1409
|
+
Use structs as data ptr.
|
1410
|
+
[bb3ce59c6acd]
|
1411
|
+
|
1412
|
+
* ext/pg_connection.c:
|
1413
|
+
Remove rb_obj_is_kind_of(connection) for every Connection method
|
1414
|
+
call.
|
1415
|
+
|
1416
|
+
There should'nt be any possibility to change the base type from Ruby
|
1417
|
+
space.
|
1418
|
+
[5dd2fec1a494]
|
1419
|
+
|
1420
|
+
* ext/pg_type_map_by_oid.c, spec/pg/type_map_by_oid_spec.rb:
|
1421
|
+
Add a OID lookup cache to PG::TypeMapByOid to the cost of 8KByte
|
1422
|
+
heap memory per instance.
|
1423
|
+
|
1424
|
+
max_rows_for_online_lookup is also increased but that's only
|
1425
|
+
guessed. So no tests were done to find any break-even point.
|
1426
|
+
[bf44b0aaf366]
|
1427
|
+
|
1428
|
+
2014-09-16 Lars Kanis <lars@greiz-reinsdorf.de>
|
1429
|
+
|
1430
|
+
* ext/pg_type_map_by_oid.c, spec/pg/type_map_by_oid_spec.rb:
|
1431
|
+
TypeMapByOid: Add threshold for per-row hash lookups.
|
1432
|
+
[892253614bdb]
|
1433
|
+
|
1434
|
+
2014-09-16 Lars Kanis <kanis@comcard.de>
|
1435
|
+
|
1436
|
+
* ext/pg_connection.c, ext/pg_result.c, ext/pg_text_decoder.c,
|
1437
|
+
lib/pg/basic_type_mapping.rb, lib/pg/result.rb,
|
1438
|
+
spec/pg/basic_type_mapping_spec.rb, spec/pg/connection_spec.rb:
|
1439
|
+
Restructured BasicTypeMapping to make use of TypeMapByOid and
|
1440
|
+
TypeMapByMriType.
|
1441
|
+
|
1442
|
+
Renamed type_map_for_result and type_map_for_query to plural form -
|
1443
|
+
that sounds better.
|
1444
|
+
[8bda62d443c9]
|
1445
|
+
|
1446
|
+
* ext/pg_type_map_by_oid.c:
|
1447
|
+
TypeMapByOid: Move init code for C-Struct to alloc function, so that
|
1448
|
+
derived classes, which don't call super in initialize don't
|
1449
|
+
segfault.
|
1450
|
+
[592fa6e19beb]
|
1451
|
+
|
1452
|
+
* ext/pg.h, ext/pg_connection.c, ext/pg_type_map_by_mri_type.c,
|
1453
|
+
spec/pg/type_map_by_mri_type_spec.rb:
|
1454
|
+
TypeMapByMriType: Allow dynamic definition of a coder for each per
|
1455
|
+
query param.
|
1456
|
+
|
1457
|
+
Remove T_NIL since it's not useful. Qnil is always mapped to SQL
|
1458
|
+
NULL.
|
1459
|
+
[8e5d18881a5d]
|
1460
|
+
|
1461
|
+
* ext/pg.h:
|
1462
|
+
Add missing function declaration.
|
1463
|
+
[30eb8599d4b2]
|
1464
|
+
|
1465
|
+
* ext/pg_result.c, ext/pg_type_map_by_column.c,
|
1466
|
+
ext/pg_type_map_by_oid.c:
|
1467
|
+
Replace some Check_Type/DATA_PTR combinations by Data_Get_Struct.
|
1468
|
+
[5f4cb85d4025]
|
1469
|
+
|
1470
|
+
* ext/pg_type_map_by_column.c:
|
1471
|
+
Remove unnecessary variable.
|
1472
|
+
[2e109387fa5b]
|
1473
|
+
|
1474
|
+
* Manifest.txt, ext/pg.c, ext/pg_type_map_by_mri_type.c,
|
1475
|
+
spec/pg/type_map_by_mri_type_spec.rb:
|
1476
|
+
Add PG::TypeMapByMriType for encoding of Ruby values.
|
1477
|
+
[00e11c3f4ced]
|
1478
|
+
|
1479
|
+
2014-09-15 Lars Kanis <lars@greiz-reinsdorf.de>
|
1480
|
+
|
1481
|
+
* ext/pg_type_map_by_column.c:
|
1482
|
+
Fix NULL pointer deref in mark function.
|
1483
|
+
|
1484
|
+
Move init of tmbc struct, so that it's done before a raise/GC run
|
1485
|
+
can happen.
|
1486
|
+
[ece47453109b]
|
1487
|
+
|
1488
|
+
2014-09-15 Lars Kanis <kanis@comcard.de>
|
1489
|
+
|
1490
|
+
* ext/pg_type_map_by_column.c:
|
1491
|
+
Avoid read of uninitialized values - detected with valgrind:
|
1492
|
+
|
1493
|
+
==30681== Conditional jump or move depends on uninitialised value(s)
|
1494
|
+
==30681== at 0xD1732EB: pg_tmbc_alloc_query_params
|
1495
|
+
(pg_type_map_by_column.c:157) ==30681== by 0x4EA93B6: rb_protect
|
1496
|
+
(eval.c:819) ==30681== by 0xD1770DE: alloc_query_params
|
1497
|
+
(pg_connection.c:1030) ==30681== by 0xD17A357: pgconn_exec_params
|
1498
|
+
(pg_connection.c:1102) ==30681== by 0x4FCE439:
|
1499
|
+
vm_call_cfunc_with_frame (vm_insnhelper.c:1470) ==30681== by
|
1500
|
+
0x4FDA48E: vm_call_method (vm_insnhelper.c:1560) ==30681== by
|
1501
|
+
0x4FD2568: vm_exec_core (insns.def:1028) ==30681== by 0x4FD7126:
|
1502
|
+
vm_exec (vm.c:1304) ==30681== by 0x4FD8EFA:
|
1503
|
+
invoke_block_from_c.part.49 (vm.c:732) ==30681== by 0x4FDCB13:
|
1504
|
+
yield_under (vm.c:704) ==30681== by 0x4FCE439:
|
1505
|
+
vm_call_cfunc_with_frame (vm_insnhelper.c:1470) ==30681== by
|
1506
|
+
0x4FDA48E: vm_call_method (vm_insnhelper.c:1560)
|
1507
|
+
[4a125325e819]
|
1508
|
+
|
1509
|
+
* ext/pg_type_map_by_column.c, ext/pg_type_map_by_oid.c:
|
1510
|
+
Fix NULL pointer dereference when a TypeMapByColumn is marked before
|
1511
|
+
#initialize
|
1512
|
+
|
1513
|
+
Remove unnecessary type cast.
|
1514
|
+
[4fe07c3ea525]
|
1515
|
+
|
1516
|
+
2014-09-14 Lars Kanis <lars@greiz-reinsdorf.de>
|
1517
|
+
|
1518
|
+
* Manifest.txt, ext/pg.c, ext/pg.h, ext/pg_type_map_by_column.c,
|
1519
|
+
ext/pg_type_map_by_oid.c, spec/pg/type_map_by_oid_spec.rb:
|
1520
|
+
Add PG::TypeMapByOid which can be used for easy and fast result
|
1521
|
+
value type casts.
|
1522
|
+
|
1523
|
+
Change TypeMapByColumn to use a GC mark function. This way we don't
|
1524
|
+
need to store the coder objects twice.
|
1525
|
+
[c967d74f8412]
|
1526
|
+
|
1527
|
+
* ext/pg_connection.c:
|
1528
|
+
Remove unnecessary memory allocation.
|
1529
|
+
[72957cc8bb04]
|
1530
|
+
|
1531
|
+
2014-09-13 Lars Kanis <lars@greiz-reinsdorf.de>
|
1532
|
+
|
1533
|
+
* spec/pg/type_map_spec.rb:
|
1534
|
+
Remove obsolete comment
|
1535
|
+
[c875975e4596]
|
1536
|
+
|
1537
|
+
* Manifest.txt, ext/pg.c, ext/pg.h, ext/pg_coder.c,
|
1538
|
+
ext/pg_column_mapping.c, ext/pg_connection.c, ext/pg_result.c,
|
1539
|
+
ext/pg_type_map.c, ext/pg_type_map_by_column.c, lib/pg.rb,
|
1540
|
+
lib/pg/basic_type_mapping.rb, lib/pg/column_mapping.rb,
|
1541
|
+
lib/pg/result.rb, lib/pg/type_map_by_column.rb,
|
1542
|
+
spec/pg/basic_type_mapping_spec.rb, spec/pg/column_mapping_spec.rb,
|
1543
|
+
spec/pg/connection_spec.rb, spec/pg/result_spec.rb,
|
1544
|
+
spec/pg/type_map_by_column_spec.rb, spec/pg/type_map_spec.rb:
|
1545
|
+
Allow different type mapping strategies beside ColumnMapping. This
|
1546
|
+
gives us the freedom to implement TypeMapByOid, TypeMapByClass or
|
1547
|
+
some fixed function mappings.
|
1548
|
+
|
1549
|
+
Rename ColumnMapping to TypeMapByColumn.
|
1550
|
+
|
1551
|
+
Seperate type_map_for_result from type_map_for_query, so thats it's
|
1552
|
+
easy to do type casts for one direction only.
|
1553
|
+
[2548e7a85049]
|
1554
|
+
|
1555
|
+
2014-09-02 Lars Kanis <lars@greiz-reinsdorf.de>
|
1556
|
+
|
1557
|
+
* ext/pg_coder.c, ext/pg_text_encoder.c, lib/pg/text_encoder.rb:
|
1558
|
+
Add documentation to some TextEncoder classes.
|
1559
|
+
[6c3fad43d3a2]
|
1560
|
+
|
1561
|
+
* ext/pg_text_encoder.c, spec/pg/type_spec.rb:
|
1562
|
+
Fix PG::TextEncoder::Identifier for more than 2 elements.
|
1563
|
+
[fcb8138f3f6f]
|
1564
|
+
|
1565
|
+
2014-08-31 Lars Kanis <lars@greiz-reinsdorf.de>
|
1566
|
+
|
1567
|
+
* ext/pg.h, ext/pg_coder.c, ext/pg_text_decoder.c,
|
1568
|
+
ext/pg_text_encoder.c, lib/pg/coder.rb, spec/pg/type_spec.rb:
|
1569
|
+
Add delimiter attribute to composite types.
|
1570
|
+
[5604eba62d48]
|
1571
|
+
|
1572
|
+
* README.rdoc:
|
1573
|
+
Update the README to reflect the API change of the type cast
|
1574
|
+
classes.
|
1575
|
+
[2afd944eadb2]
|
1576
|
+
|
1577
|
+
* lib/pg/basic_type_mapping.rb:
|
1578
|
+
Update PG::BasicTypeMapping to make use of new PG::Coder classes.
|
1579
|
+
Add some documentation to BasicTypeMapping.
|
1580
|
+
[ad24be2307d1]
|
1581
|
+
|
1582
|
+
* spec/pg/type_spec.rb:
|
1583
|
+
type_spec doesn't make use of the helpers.
|
1584
|
+
[7a3a4197a61f]
|
1585
|
+
|
1586
|
+
2014-08-26 Lars Kanis <lars@greiz-reinsdorf.de>
|
1587
|
+
|
1588
|
+
* Manifest.txt, ext/pg.c, ext/pg.h, ext/pg_binary_decoder.c,
|
1589
|
+
ext/pg_binary_encoder.c, ext/pg_coder.c, ext/pg_column_mapping.c,
|
1590
|
+
ext/pg_connection.c, ext/pg_text_decoder.c, ext/pg_text_encoder.c,
|
1591
|
+
ext/pg_type.c, lib/pg.rb, lib/pg/basic_type_mapping.rb,
|
1592
|
+
lib/pg/coder.rb, lib/pg/column_mapping.rb, lib/pg/text_decoder.rb,
|
1593
|
+
lib/pg/text_encoder.rb, lib/pg/type.rb,
|
1594
|
+
spec/pg/column_mapping_spec.rb, spec/pg/result_spec.rb,
|
1595
|
+
spec/pg/type_spec.rb:
|
1596
|
+
Simplify the type cast API by merging PG::Type and PG::Coder
|
1597
|
+
classes.
|
1598
|
+
|
1599
|
+
En/decoders can now be built directly as
|
1600
|
+
PG::TextDecoder::Integer.new instead of previous
|
1601
|
+
PG::Type.new(decoder: PG::TextDecoder::Integer) .
|
1602
|
+
[2b428d490e08]
|
1603
|
+
|
1604
|
+
2014-08-24 Lars Kanis <lars@greiz-reinsdorf.de>
|
1605
|
+
|
1606
|
+
* ext/pg_connection.c, ext/pg_type.c:
|
1607
|
+
Change all remaining Oid <-> VALUE conversions to UINT.
|
1608
|
+
[66047d8245d1]
|
1609
|
+
|
1610
|
+
* ext/pg_coder.c:
|
1611
|
+
Remove unused accessors.
|
1612
|
+
[b6b84993c6ff]
|
1613
|
+
|
1614
|
+
* spec/pg/type_spec.rb:
|
1615
|
+
Convert remaining two should's to expect syntax.
|
1616
|
+
[1888e8425d07]
|
1617
|
+
|
1618
|
+
* .travis.yml:
|
1619
|
+
'gem install rake-compiler' is still needed for cross build.
|
1620
|
+
[f8a0ac84f72f]
|
1621
|
+
|
1622
|
+
* .travis.yml, spec/pg/connection_spec.rb:
|
1623
|
+
Merge branch 'master' into column_type_mapping
|
1624
|
+
|
1625
|
+
Conflicts: .travis.yml
|
1626
|
+
[36684ebbc21d]
|
1627
|
+
|
1628
|
+
2014-08-22 Lars Kanis <lars@greiz-reinsdorf.de>
|
1629
|
+
|
1630
|
+
* Merge /home/lars/comcard/tmp5/ruby-pg.git.bare into
|
1631
|
+
column_type_mapping
|
1632
|
+
[2485600065c4]
|
1633
|
+
|
1634
|
+
* spec/helpers.rb, spec/pg/column_mapping_spec.rb,
|
1635
|
+
spec/pg/connection_spec.rb, spec/pg/result_spec.rb,
|
1636
|
+
spec/pg/type_spec.rb:
|
1637
|
+
Convert type cast specs to expect syntax of RSpec 3
|
1638
|
+
[3497aade8d9a]
|
1639
|
+
|
1640
|
+
* Manifest.txt, ext/pg_connection.c, ext/pg_result.c, spec/helpers.rb,
|
1641
|
+
spec/lib/helpers.rb, spec/pg/connection_spec.rb,
|
1642
|
+
spec/pg/result_spec.rb:
|
1643
|
+
Merge branch 'master' into column_type_mapping
|
1644
|
+
|
1645
|
+
Conflicts: Manifest.txt spec/helpers.rb spec/pg/connection_spec.rb
|
1646
|
+
spec/pg/result_spec.rb
|
1647
|
+
[8eb15150e2f7]
|
1648
|
+
|
1649
|
+
2014-05-08 Lars Kanis <kanis@comcard.de>
|
1650
|
+
|
1651
|
+
* lib/pg/basic_type_mapping.rb:
|
1652
|
+
Make column_mapping_for_* methods public.
|
1653
|
+
[5b2c4584de20]
|
1654
|
+
|
1655
|
+
2014-04-27 Lars Kanis <lars@greiz-reinsdorf.de>
|
1656
|
+
|
1657
|
+
* lib/pg/coder.rb, lib/pg/type.rb:
|
1658
|
+
Add nicer inspect methods to Type and Coder.
|
1659
|
+
[bea1b87d449f]
|
1660
|
+
|
1661
|
+
2014-04-23 Lars Kanis <lars@greiz-reinsdorf.de>
|
1662
|
+
|
1663
|
+
* ext/pg_text_encoder.c, spec/pg/type_spec.rb:
|
1664
|
+
Add PG::TextEncoder::QUOTED_LITERAL.
|
1665
|
+
|
1666
|
+
This can be used to encode values directly for sql insertion.
|
1667
|
+
[899ae14a4177]
|
1668
|
+
|
1669
|
+
* ext/pg.h, ext/pg_binary_encoder.c, ext/pg_text_encoder.c,
|
1670
|
+
spec/pg/type_spec.rb:
|
1671
|
+
Use #to_i for encoding of non Numeric types. That allows use of
|
1672
|
+
strings like "123-garbage" as integer input.
|
1673
|
+
[3771cd7756af]
|
1674
|
+
|
1675
|
+
* ext/pg_text_encoder.c, lib/pg/basic_type_mapping.rb,
|
1676
|
+
spec/pg/column_mapping_spec.rb:
|
1677
|
+
Remove PG::TextEncoder::BYTEA, as it was not correct.
|
1678
|
+
[f246df175b01]
|
1679
|
+
|
1680
|
+
2014-04-22 Lars Kanis <lars@greiz-reinsdorf.de>
|
1681
|
+
|
1682
|
+
* ext/pg_connection.c, spec/pg/column_mapping_spec.rb:
|
1683
|
+
Fix encoding of NULL values to typed fields.
|
1684
|
+
[8820a9c4271e]
|
1685
|
+
|
1686
|
+
* ext/pg_binary_encoder.c, ext/pg_text_encoder.c,
|
1687
|
+
spec/pg/type_spec.rb:
|
1688
|
+
Allow to encode Strings to int types
|
1689
|
+
[b8c6be80a931]
|
1690
|
+
|
1691
|
+
* ext/pg_text_encoder.c, spec/pg/type_spec.rb:
|
1692
|
+
Fix encoding of empty arrays.
|
1693
|
+
[eff38fc911d0]
|
1694
|
+
|
1695
|
+
* ext/pg_binary_decoder.c, ext/pg_binary_encoder.c,
|
1696
|
+
ext/pg_text_decoder.c, ext/pg_text_encoder.c,
|
1697
|
+
lib/pg/basic_type_mapping.rb, lib/pg/text_decoder.rb,
|
1698
|
+
lib/pg/text_encoder.rb, spec/pg/column_mapping_spec.rb,
|
1699
|
+
spec/pg/result_spec.rb, spec/pg/type_spec.rb:
|
1700
|
+
Change all coder objects to UPPERCASE, because they aren't class
|
1701
|
+
constants.
|
1702
|
+
[8384fe858ebf]
|
1703
|
+
|
1704
|
+
2014-04-21 Lars Kanis <lars@greiz-reinsdorf.de>
|
1705
|
+
|
1706
|
+
* ext/pg_text_decoder.c, ext/pg_text_encoder.c, spec/pg/type_spec.rb:
|
1707
|
+
Add encoder and decoder for identifiers.
|
1708
|
+
|
1709
|
+
Improve quotation code of Array encoder.
|
1710
|
+
|
1711
|
+
Refactor Array encoder and decoder functions to allow reuse for
|
1712
|
+
identifiers.
|
1713
|
+
[b13e27b316d4]
|
1714
|
+
|
1715
|
+
2014-04-19 Lars Kanis <lars@greiz-reinsdorf.de>
|
1716
|
+
|
1717
|
+
* ext/pg_type.c, spec/pg/type_spec.rb:
|
1718
|
+
Allow PG::CompositeType#elements_type=nil and fix error text for
|
1719
|
+
invalid types.
|
1720
|
+
[91f8becf18aa]
|
1721
|
+
|
1722
|
+
* spec/pg/column_mapping_spec.rb:
|
1723
|
+
Add decoder tests for float values NAN and INFINITY.
|
1724
|
+
[c45b87b317c3]
|
1725
|
+
|
1726
|
+
* ext/pg_type.c:
|
1727
|
+
pg_type_get_and_check() is no longer in use.
|
1728
|
+
[f387ba78ccbf]
|
1729
|
+
|
1730
|
+
* lib/pg/basic_type_mapping.rb:
|
1731
|
+
Array.new is somewhat faster than using an Enumerator.
|
1732
|
+
[458cdea12837]
|
1733
|
+
|
1734
|
+
* ext/pg_column_mapping.c, ext/pg_result.c:
|
1735
|
+
Don't call ENCODING_GET() for each result value, but only once, when
|
1736
|
+
the ColumnMapping is assigned to a PG::Result.
|
1737
|
+
|
1738
|
+
The speedup is only minimal (~1%).
|
1739
|
+
[29fc4769e95a]
|
1740
|
+
|
1741
|
+
2014-04-18 Lars Kanis <lars@greiz-reinsdorf.de>
|
1742
|
+
|
1743
|
+
* ext/pg_type.c:
|
1744
|
+
Add check for output lengths of first/second pass of C based
|
1745
|
+
encoders.
|
1746
|
+
[445a04470387]
|
1747
|
+
|
1748
|
+
* ext/pg_text_encoder.c, spec/pg/type_spec.rb:
|
1749
|
+
Fix size computation of negative numbers for text integer encoding.
|
1750
|
+
[3984ee009617]
|
1751
|
+
|
1752
|
+
* README.rdoc:
|
1753
|
+
Tune type cast documentation a bit more.
|
1754
|
+
[65f0f57deb57]
|
1755
|
+
|
1756
|
+
* README.rdoc:
|
1757
|
+
Add the most simple use case for type casting to the README.
|
1758
|
+
[6524afce6fbc]
|
1759
|
+
|
1760
|
+
* ext/pg_text_decoder.c:
|
1761
|
+
rb_cstr2inum() seems to be slow, so we do the integer conversion by
|
1762
|
+
hand in simple cases. This proved to be 40% faster.
|
1763
|
+
[976fcccb023f]
|
1764
|
+
|
1765
|
+
2014-04-16 Lars Kanis <lars@greiz-reinsdorf.de>
|
1766
|
+
|
1767
|
+
* spec/pg/type_spec.rb:
|
1768
|
+
Add spec for PG::Type default values.
|
1769
|
+
[ac8037456424]
|
1770
|
+
|
1771
|
+
* lib/pg/type.rb:
|
1772
|
+
Remove old code.
|
1773
|
+
[d9abf3b53985]
|
1774
|
+
|
1775
|
+
* spec/pg/type_spec.rb:
|
1776
|
+
Add specs for Type#to_h.
|
1777
|
+
[170f9f5a344b]
|
1778
|
+
|
1779
|
+
* ext/pg_type.c:
|
1780
|
+
needs_quotation? should return true or false.
|
1781
|
+
[f00c1948e602]
|
1782
|
+
|
1783
|
+
2014-04-16 Lars Kanis <kanis@comcard.de>
|
1784
|
+
|
1785
|
+
* History.rdoc, lib/pg.rb:
|
1786
|
+
Bump VERSION to 0.18.0
|
1787
|
+
[91ea8734d053]
|
1788
|
+
|
1789
|
+
2014-04-15 Lars Kanis <lars@greiz-reinsdorf.de>
|
1790
|
+
|
1791
|
+
* ext/pg.h, ext/pg_binary_decoder.c, ext/pg_binary_encoder.c,
|
1792
|
+
ext/pg_coder.c, ext/pg_text_decoder.c, ext/pg_text_encoder.c,
|
1793
|
+
lib/pg/coder.rb:
|
1794
|
+
Use full qualified name for Coder#name.
|
1795
|
+
|
1796
|
+
Remove duplicated code in coder definitions.
|
1797
|
+
[7ca87a46a91c]
|
1798
|
+
|
1799
|
+
2014-04-15 Lars Kanis <kanis@comcard.de>
|
1800
|
+
|
1801
|
+
* Manifest.txt:
|
1802
|
+
Update Manifest.txt
|
1803
|
+
[02c6ce33a90c]
|
1804
|
+
|
1805
|
+
* ext/pg_binary_decoder.c, ext/pg_binary_encoder.c,
|
1806
|
+
ext/pg_text_decoder.c, ext/pg_text_encoder.c:
|
1807
|
+
Remove duplicated @name accessors.
|
1808
|
+
[13cc22a2b729]
|
1809
|
+
|
1810
|
+
2014-04-15 Lars Kanis <lars@greiz-reinsdorf.de>
|
1811
|
+
|
1812
|
+
* ext/pg.c, ext/pg.h, ext/pg_binary_decoder.c,
|
1813
|
+
ext/pg_binary_encoder.c, ext/pg_coder.c, ext/pg_text_decoder.c,
|
1814
|
+
ext/pg_text_encoder.c, lib/pg.rb, lib/pg/coder.rb, lib/pg/type.rb,
|
1815
|
+
spec/pg/type_spec.rb:
|
1816
|
+
Add ability to Marshal.dump/load PG::Type objects.
|
1817
|
+
|
1818
|
+
This should help to use PG::Type in ActiveRecord.
|
1819
|
+
|
1820
|
+
This also fixes base classes for Composite coders.
|
1821
|
+
[ef0ef4b51e0d]
|
1822
|
+
|
1823
|
+
2014-04-14 Lars Kanis <kanis@comcard.de>
|
1824
|
+
|
1825
|
+
* README.rdoc:
|
1826
|
+
Update README according to changes in type mapping.
|
1827
|
+
[fe3b1cd9850b]
|
1828
|
+
|
1829
|
+
2014-04-13 Lars Kanis <lars@greiz-reinsdorf.de>
|
1830
|
+
|
1831
|
+
* Manifest.txt, ext/pg.h, ext/pg_column_mapping.c, ext/pg_type.c,
|
1832
|
+
lib/pg.rb, lib/pg/simple_type.rb, lib/pg/type.rb,
|
1833
|
+
spec/pg/type_spec.rb:
|
1834
|
+
Add a common base class PG::Type for PG::SimpleType and
|
1835
|
+
PG::CompositeType.
|
1836
|
+
|
1837
|
+
"some kind of PG::Type" sounds a little bit better than SimpleType.
|
1838
|
+
[2a385bd22f7c]
|
1839
|
+
|
1840
|
+
* Manifest.txt, ext/pg.c, ext/pg.h, ext/pg_binary_decoder.c,
|
1841
|
+
ext/pg_binary_encoder.c, ext/pg_column_mapping.c,
|
1842
|
+
ext/pg_text_decoder.c, ext/pg_text_encoder.c, ext/pg_type.c,
|
1843
|
+
ext/pg_type_binary_decoder.c, ext/pg_type_binary_encoder.c,
|
1844
|
+
ext/pg_type_text_decoder.c, ext/pg_type_text_encoder.c, lib/pg.rb,
|
1845
|
+
lib/pg/basic_type_mapping.rb, lib/pg/simple_type.rb,
|
1846
|
+
lib/pg/text_decoder.rb, lib/pg/text_encoder.rb,
|
1847
|
+
lib/pg/type/simple_type.rb, lib/pg/type/text_decoder.rb,
|
1848
|
+
lib/pg/type/text_encoder.rb, lib/pg/type/type.rb,
|
1849
|
+
spec/pg/column_mapping_spec.rb, spec/pg/result_spec.rb,
|
1850
|
+
spec/pg/type_spec.rb:
|
1851
|
+
Remove namespace PG::Type and move all contained classes in PG
|
1852
|
+
namespace.
|
1853
|
+
|
1854
|
+
Rename function names accordingly.
|
1855
|
+
[c5886a74623c]
|
1856
|
+
|
1857
|
+
2014-04-12 Lars Kanis <lars@greiz-reinsdorf.de>
|
1858
|
+
|
1859
|
+
* spec/pg/result_spec.rb:
|
1860
|
+
Remove needless == in spec.
|
1861
|
+
[03d3a7c68f8f]
|
1862
|
+
|
1863
|
+
* Manifest.txt, lib/pg.rb, lib/pg/basic_type_mapping.rb,
|
1864
|
+
lib/pg/column_mapping.rb, lib/pg/result.rb, lib/pg/type.rb,
|
1865
|
+
spec/pg/column_mapping_spec.rb, spec/pg/connection_spec.rb,
|
1866
|
+
spec/pg/result_spec.rb:
|
1867
|
+
Change BasicTypeMapping so that it uses type information from the
|
1868
|
+
DB.
|
1869
|
+
|
1870
|
+
Align specs to the latest type code refactoring.
|
1871
|
+
[ad90adcc1ae3]
|
1872
|
+
|
1873
|
+
* ext/pg_column_mapping.c, lib/pg/type/simple_type.rb,
|
1874
|
+
lib/pg/type/text_decoder.rb, lib/pg/type/text_encoder.rb,
|
1875
|
+
lib/pg/type/type.rb:
|
1876
|
+
Fix colmap_init. It was the wrong variable name.
|
1877
|
+
[eb13e786adec]
|
1878
|
+
|
1879
|
+
2014-04-11 Lars Kanis <lars@greiz-reinsdorf.de>
|
1880
|
+
|
1881
|
+
* spec/pg/type_spec.rb:
|
1882
|
+
Convert and add several new test cases to SimpleType and
|
1883
|
+
CompositType.
|
1884
|
+
[d07abe17054c]
|
1885
|
+
|
1886
|
+
* ext/pg_type_text_encoder.c:
|
1887
|
+
Avoid calling enc_func twice for array encoding.
|
1888
|
+
|
1889
|
+
This did lead to wrong behaviour on array of ruby defined types when
|
1890
|
+
using quotation.
|
1891
|
+
[767e1049ea08]
|
1892
|
+
|
1893
|
+
* ext/pg_type_binary_decoder.c, ext/pg_type_binary_encoder.c,
|
1894
|
+
ext/pg_type_text_decoder.c, ext/pg_type_text_encoder.c:
|
1895
|
+
Add String en/decoder.
|
1896
|
+
[d420f9b92bf3]
|
1897
|
+
|
1898
|
+
* ext/pg_type.c:
|
1899
|
+
Ensure encoder output is a string and decoder input is a string.
|
1900
|
+
[fd943340acac]
|
1901
|
+
|
1902
|
+
2014-04-09 Lars Kanis <lars@greiz-reinsdorf.de>
|
1903
|
+
|
1904
|
+
* ext/pg.c, ext/pg.h, ext/pg_column_mapping.c, ext/pg_connection.c,
|
1905
|
+
ext/pg_type.c, ext/pg_type_binary_decoder.c,
|
1906
|
+
ext/pg_type_binary_encoder.c, ext/pg_type_text_decoder.c,
|
1907
|
+
ext/pg_type_text_encoder.c:
|
1908
|
+
WIP: Major refactoring of the type mapping code and API.
|
1909
|
+
|
1910
|
+
Now we distinguish between Encoders/Decoders and Types on Ruby
|
1911
|
+
level.
|
1912
|
+
[48e40627d66b]
|
1913
|
+
|
1914
|
+
2014-04-04 Lars Kanis <lars@greiz-reinsdorf.de>
|
1915
|
+
|
1916
|
+
* spec/pg/connection_spec.rb:
|
1917
|
+
Cleanup test tables, so that these two test cases can be executed
|
1918
|
+
multiple times.
|
1919
|
+
[3c450594f574]
|
1920
|
+
|
1921
|
+
2014-03-31 Lars Kanis <kanis@comcard.de>
|
1922
|
+
|
1923
|
+
* ext/pg.h, ext/pg_column_mapping.c, ext/pg_type.c, lib/pg/type.rb,
|
1924
|
+
spec/pg/type_spec.rb:
|
1925
|
+
Allow composition of types in Ruby space, so that we can combine
|
1926
|
+
Ruby and C converters.
|
1927
|
+
[e25ee59f5bff]
|
1928
|
+
|
1929
|
+
2014-03-31 Lars Kanis <lars@greiz-reinsdorf.de>
|
1930
|
+
|
1931
|
+
* ext/pg.h, ext/pg_column_mapping.c, ext/pg_connection.c,
|
1932
|
+
ext/pg_type.c:
|
1933
|
+
Add composit types that allow mixed C/Ruby implementation of
|
1934
|
+
base/elem converters.
|
1935
|
+
[c5a3c68fe2bb]
|
1936
|
+
|
1937
|
+
2014-03-24 Lars Kanis <kanis@comcard.de>
|
1938
|
+
|
1939
|
+
* .travis.yml:
|
1940
|
+
Fix windows build on travis-ci.
|
1941
|
+
[0fc7eee486c8]
|
1942
|
+
|
1943
|
+
* lib/pg/type_mapping.rb:
|
1944
|
+
Merge branch 'column_type_mapping' of https://github.com/larskanis
|
1945
|
+
/ruby-pg into column_type_mapping
|
1946
|
+
[8e785f697ba7]
|
1947
|
+
|
1948
|
+
2014-03-23 Lars Kanis <lars@greiz-reinsdorf.de>
|
1949
|
+
|
1950
|
+
* Manifest.txt, spec/pg/array_parser_spec.rb, spec/pg/type_spec.rb:
|
1951
|
+
Merge array_parser_spec.rb into type_spec.rb. Update Manifest.txt.
|
1952
|
+
[23907cc8d0f2]
|
1953
|
+
|
1954
|
+
* spec/lib/helpers.rb, spec/pg/column_mapping_spec.rb:
|
1955
|
+
Type encoding: Verify that the type OIDs match our expectation when
|
1956
|
+
encoding values.
|
1957
|
+
[4a4f58d3a058]
|
1958
|
+
|
1959
|
+
2014-03-22 Lars Kanis <lars@greiz-reinsdorf.de>
|
1960
|
+
|
1961
|
+
* ext/pg_type.c, spec/pg/type_spec.rb:
|
1962
|
+
Fix conversion of Bignums to integer for text format.
|
1963
|
+
[987713eeca05]
|
1964
|
+
|
1965
|
+
* lib/pg/basic_type_mapping.rb, spec/pg/array_parser_spec.rb,
|
1966
|
+
spec/pg/column_mapping_spec.rb:
|
1967
|
+
Add basic type mapping for array types.
|
1968
|
+
[112cc249775b]
|
1969
|
+
|
1970
|
+
* ext/pg_type.c:
|
1971
|
+
Add float type conversions for text format.
|
1972
|
+
[d45c09db59c9]
|
1973
|
+
|
1974
|
+
* ext/pg_type.c:
|
1975
|
+
Text array encoding: Store all intermediate values into a single
|
1976
|
+
array even of converting multi dimensional arrays.
|
1977
|
+
|
1978
|
+
Add integer encoding for text format.
|
1979
|
+
[d95cc34a5119]
|
1980
|
+
|
1981
|
+
* ext/pg_type.c:
|
1982
|
+
Add inttypes.h since we're using int_t types.
|
1983
|
+
[c3a4655c70c5]
|
1984
|
+
|
1985
|
+
* ext/pg_type.c:
|
1986
|
+
Array type encoding: Avoid unnecessary copying of data.
|
1987
|
+
[78a8caa47042]
|
1988
|
+
|
1989
|
+
2014-03-21 Lars Kanis <lars@greiz-reinsdorf.de>
|
1990
|
+
|
1991
|
+
* ext/pg_type.c, spec/pg/array_parser_spec.rb:
|
1992
|
+
Add array encoding for text format.
|
1993
|
+
[54682c69398f]
|
1994
|
+
|
1995
|
+
* spec/pg/column_mapping_spec.rb:
|
1996
|
+
Fix typo in column_mapping_spec.rb
|
1997
|
+
[007d45ed9ab5]
|
1998
|
+
|
1999
|
+
* ext/pg_column_mapping.c:
|
2000
|
+
Fix wrong type check due to 274d3e533.
|
2001
|
+
[51a9632e9a33]
|
2002
|
+
|
2003
|
+
2014-02-25 Lars Kanis <lars@greiz-reinsdorf.de>
|
2004
|
+
|
2005
|
+
* ext/pg_type.c:
|
2006
|
+
Add encode/decode methods to CConverters only if corresponding C
|
2007
|
+
functions are defined.
|
2008
|
+
|
2009
|
+
This allowes simple detection of the capablities per respond_to? .
|
2010
|
+
[16209f4f9b70]
|
2011
|
+
|
2012
|
+
* ext/pg_connection.c:
|
2013
|
+
Use variable length arrays on the stack for intermediates while
|
2014
|
+
param value convertion.
|
2015
|
+
|
2016
|
+
This avoids malloc/free overhead, frees the memory before the
|
2017
|
+
function call and ensures proper GC protection.
|
2018
|
+
[44a016bc84fa]
|
2019
|
+
|
2020
|
+
2014-02-19 Lars Kanis <lars@greiz-reinsdorf.de>
|
2021
|
+
|
2022
|
+
* ext/pg_type.c:
|
2023
|
+
Fix value.inspect in raised exception.
|
2024
|
+
[2af869fcf933]
|
2025
|
+
|
2026
|
+
* README.rdoc:
|
2027
|
+
Add paths to implementation files in README.
|
2028
|
+
[d6ca799ef38c]
|
2029
|
+
|
2030
|
+
2014-02-18 Lars Kanis <lars@greiz-reinsdorf.de>
|
2031
|
+
|
2032
|
+
* ext/pg_result.c:
|
2033
|
+
Avoid 'result has been cleared' error in case of UnableToSend.
|
2034
|
+
|
2035
|
+
Avoid double checking per pgresult_get().
|
2036
|
+
[4a478da0839a]
|
2037
|
+
|
2038
|
+
* README.rdoc, ext/pg_connection.c, ext/pg_result.c,
|
2039
|
+
spec/pg/connection_spec.rb:
|
2040
|
+
Allow connection wide assignment of type_mapping.
|
2041
|
+
[ecd58ff267b1]
|
2042
|
+
|
2043
|
+
* ext/pg_type.c:
|
2044
|
+
Add missing string termination to array conversion.
|
2045
|
+
[f2dcc3692737]
|
2046
|
+
|
2047
|
+
* ext/pg_type.c, spec/pg/array_parser_spec.rb,
|
2048
|
+
spec/pg/column_mapping_spec.rb:
|
2049
|
+
Add array type conversions.
|
2050
|
+
[e667a761e93b]
|
2051
|
+
|
2052
|
+
* ext/pg_type.c, spec/pg/type_spec.rb:
|
2053
|
+
Allow Type.decode to be called without tuple and field args.
|
2054
|
+
[e6089af0d283]
|
2055
|
+
|
2056
|
+
* lib/pg.rb, lib/pg/basic_type_mapping.rb, lib/pg/type_mapping.rb:
|
2057
|
+
Rename file so that it matches the class name.
|
2058
|
+
[95a4b6841516]
|
2059
|
+
|
2060
|
+
2014-02-13 Lars Kanis <lars@greiz-reinsdorf.de>
|
2061
|
+
|
2062
|
+
* .travis.yml:
|
2063
|
+
Update travis config.
|
2064
|
+
[8082f9d17c76]
|
2065
|
+
|
2066
|
+
* README.rdoc, lib/pg/result.rb:
|
2067
|
+
Tweak documentation.
|
2068
|
+
[c5e439d9ee9d]
|
2069
|
+
|
2070
|
+
* spec/pg/column_mapping_spec.rb:
|
2071
|
+
Fix test for time input type cast.
|
2072
|
+
[a217e1a4d4ac]
|
2073
|
+
|
2074
|
+
* spec/pg/connection_spec.rb:
|
2075
|
+
Do raise test per plpgsql on PostgreSQL-9.2+ only.
|
2076
|
+
[99301f365b41]
|
2077
|
+
|
2078
|
+
2014-02-13 Lars Kanis <kanis@comcard.de>
|
2079
|
+
|
2080
|
+
* spec/pg/connection_spec.rb:
|
2081
|
+
Restrict raise test with plpgsql to PostgreSQL-9.0+
|
2082
|
+
[62ad91254117]
|
2083
|
+
|
2084
|
+
2014-02-12 Lars Kanis <lars@greiz-reinsdorf.de>
|
2085
|
+
|
2086
|
+
* lib/pg/type_mapping.rb:
|
2087
|
+
Add missing type_mapping.rb
|
2088
|
+
[33238fce4d0b]
|
2089
|
+
|
2090
|
+
* README.rdoc, ext/pg_connection.c, ext/pg_result.c, lib/pg.rb,
|
2091
|
+
lib/pg/column_mapping.rb, lib/pg/result.rb, lib/pg/type.rb,
|
2092
|
+
spec/pg/column_mapping_spec.rb, spec/pg/connection_spec.rb,
|
2093
|
+
spec/pg/result_spec.rb:
|
2094
|
+
Implement BasicTypeMapping as a replacement to the Hash based
|
2095
|
+
DEFAULT_OID/DEFAULT_TYPE mechanism.
|
2096
|
+
[08fe2daeda55]
|
2097
|
+
|
2098
|
+
2014-02-11 Lars Kanis <lars@greiz-reinsdorf.de>
|
2099
|
+
|
2100
|
+
* README.rdoc:
|
2101
|
+
Add description of the type cast architecture.
|
2102
|
+
[e57f8662f564]
|
2103
|
+
|
2104
|
+
2014-01-06 Lars Kanis <lars@greiz-reinsdorf.de>
|
2105
|
+
|
2106
|
+
* ext/pg_connection.c, spec/pg/connection_spec.rb:
|
2107
|
+
Ensure all memory is free'd in case of exceptions in
|
2108
|
+
alloc_query_params()
|
2109
|
+
[efb0fdf2af3d]
|
2110
|
+
|
2111
|
+
2014-01-01 Lars Kanis <lars@greiz-reinsdorf.de>
|
2112
|
+
|
2113
|
+
* ext/pg.h, ext/pg_column_mapping.c, ext/pg_connection.c,
|
2114
|
+
ext/pg_type.c, lib/pg.rb, lib/pg/column_mapping.rb,
|
2115
|
+
lib/pg/result.rb, lib/pg/type.rb, spec/pg/column_mapping_spec.rb,
|
2116
|
+
spec/pg/result_spec.rb:
|
2117
|
+
Add optional type casting to input parameters
|
2118
|
+
|
2119
|
+
Refactor decoding code
|
2120
|
+
[19491546392d]
|
2121
|
+
|
2122
|
+
2013-12-28 Lars Kanis <lars@greiz-reinsdorf.de>
|
2123
|
+
|
2124
|
+
* ext/pg_connection.c:
|
2125
|
+
Refactor processing of query params, which was almost equal for
|
2126
|
+
exec_params, exec_prepared, send_query and send_query_prepared.
|
2127
|
+
[a3b5e9405825]
|
2128
|
+
|
2129
|
+
2013-12-23 Lars Kanis <lars@greiz-reinsdorf.de>
|
2130
|
+
|
2131
|
+
* ext/pg.h, ext/pg_column_mapping.c, ext/pg_type.c, lib/pg/type.rb,
|
2132
|
+
spec/pg/column_mapping_spec.rb, spec/pg/type_spec.rb:
|
2133
|
+
Detach type decoders from Result, so that they could be used with
|
2134
|
+
COPY.
|
2135
|
+
|
2136
|
+
Change type encoders to a two pass mechanism, which could also be
|
2137
|
+
used with COPY.
|
2138
|
+
[40e8c382c3ad]
|
2139
|
+
|
2140
|
+
2013-12-21 Lars Kanis <lars@greiz-reinsdorf.de>
|
2141
|
+
|
2142
|
+
* ext/pg_column_mapping.c, ext/pg_type.c:
|
2143
|
+
Always map NULL values to Qnil.
|
2144
|
+
|
2145
|
+
This removes duplicated code from type converters and will speedup
|
2146
|
+
converters implemented in ruby.
|
2147
|
+
[af270f5e3ddf]
|
2148
|
+
|
2149
|
+
2013-12-13 Lars Kanis <lars@greiz-reinsdorf.de>
|
2150
|
+
|
2151
|
+
* ext/pg.h, ext/pg_column_mapping.c, ext/pg_type.c, lib/pg/result.rb,
|
2152
|
+
lib/pg/type.rb, spec/pg/column_mapping_spec.rb,
|
2153
|
+
spec/pg/result_spec.rb, spec/pg/type_spec.rb:
|
2154
|
+
more WIP: Add bidirectional type mapping
|
2155
|
+
[1ea69ac572c0]
|
2156
|
+
|
2157
|
+
2013-12-09 Lars Kanis <lars@greiz-reinsdorf.de>
|
2158
|
+
|
2159
|
+
* Manifest.txt, ext/pg.c, ext/pg.h, ext/pg_column_mapping.c,
|
2160
|
+
ext/pg_result.c, ext/pg_type.c, lib/pg.rb, lib/pg/result.rb,
|
2161
|
+
lib/pg/type.rb, spec/pg/column_mapping_spec.rb,
|
2162
|
+
spec/pg/result_spec.rb, spec/pg/type_spec.rb:
|
2163
|
+
WIP: Add bidiractional type mapping
|
2164
|
+
[a89d4991f444]
|
2165
|
+
|
2166
|
+
2013-12-07 Lars Kanis <lars@greiz-reinsdorf.de>
|
2167
|
+
|
2168
|
+
* lib/pg/result.rb, spec/pg/column_mapping_spec.rb:
|
2169
|
+
Add a (quite incomplete) default mapping of OIDs to PG::Result and
|
2170
|
+
allow distinct mappings for text and binary formats to
|
2171
|
+
Result#map_types! .
|
2172
|
+
|
2173
|
+
Add documentation to Result#map_types! .
|
2174
|
+
[d137dfe12905]
|
2175
|
+
|
2176
|
+
2013-07-19 Lars Kanis <kanis@comcard.de>
|
2177
|
+
|
2178
|
+
* spec/pg/column_mapping_spec.rb:
|
2179
|
+
Too less quoting. Should fix compat with PostgreSQL 8.4.
|
2180
|
+
[f502d245ea55]
|
2181
|
+
|
2182
|
+
2013-07-18 Lars Kanis <kanis@comcard.de>
|
2183
|
+
|
2184
|
+
* Manifest.txt:
|
2185
|
+
Add missing files to Manifest.txt.
|
2186
|
+
[b5c94ff41d4e]
|
2187
|
+
|
2188
|
+
* Manifest.txt, ext/pg_column_mapping.c, ext/util.c, ext/util.h:
|
2189
|
+
Add compat with Windows platform.
|
2190
|
+
[5780ddc9c111]
|
2191
|
+
|
2192
|
+
* ext/pg_column_mapping.c, spec/pg/column_mapping_spec.rb:
|
2193
|
+
Add compat with Ruby-1.8.7.
|
2194
|
+
[ab892e597042]
|
2195
|
+
|
2196
|
+
* spec/pg/column_mapping_spec.rb:
|
2197
|
+
Use old BYTEA escaping for compat with PostgreSQL 8.4
|
2198
|
+
[aa7b76abda66]
|
2199
|
+
|
2200
|
+
2013-07-04 Lars Kanis <lars@greiz-reinsdorf.de>
|
2201
|
+
|
2202
|
+
* ext/pg_column_mapping.c, spec/pg/column_mapping_spec.rb:
|
2203
|
+
Add BinaryBolean converter.
|
2204
|
+
|
2205
|
+
Refactor converter tests.
|
2206
|
+
[7f70efce09f0]
|
2207
|
+
|
2208
|
+
* ext/pg_column_mapping.c, spec/pg/column_mapping_spec.rb:
|
2209
|
+
Add BinaryFloat converter.
|
2210
|
+
[57942116c746]
|
2211
|
+
|
2212
|
+
2013-07-03 Lars Kanis <lars@greiz-reinsdorf.de>
|
2213
|
+
|
2214
|
+
* ext/pg_column_mapping.c, lib/pg/result.rb,
|
2215
|
+
spec/pg/column_mapping_spec.rb, spec/pg/result_spec.rb:
|
2216
|
+
Change ColumnMapping#initialize to take one Array arg rather than a
|
2217
|
+
parameter list.
|
2218
|
+
|
2219
|
+
Suggested by Aaron: https://groups.google.com/d/msg/ruby-
|
2220
|
+
pg/PE5K6q9RzBs/M5prOno39DEJ
|
2221
|
+
[2606fc4f350e]
|
2222
|
+
|
2223
|
+
* spec/pg/result_spec.rb:
|
2224
|
+
Add test case for all value retrieving methods of PG::Result .
|
2225
|
+
[970d02cee164]
|
2226
|
+
|
2227
|
+
* spec/pg/column_mapping_spec.rb:
|
2228
|
+
Add test case for default_mapping in Result#map_types!
|
2229
|
+
[dbcf59dcc118]
|
2230
|
+
|
2231
|
+
* lib/pg/result.rb, spec/pg/column_mapping_spec.rb:
|
2232
|
+
Add convenience method PG::Result#map_types!
|
2233
|
+
[3e900bc457e8]
|
2234
|
+
|
2235
|
+
* spec/pg/column_mapping_spec.rb:
|
2236
|
+
Add Boolean ColumnMapping spec and do Integer a bit more compact.
|
2237
|
+
[34f6e4b23fae]
|
2238
|
+
|
2239
|
+
2013-07-01 Lars Kanis <kanis@comcard.de>
|
2240
|
+
|
2241
|
+
* ext/pg_column_mapping.c, spec/pg/column_mapping_spec.rb:
|
2242
|
+
Add converter BinaryInteger.
|
2243
|
+
[6ccf466e1537]
|
2244
|
+
|
2245
|
+
* ext/pg_column_mapping.c, spec/pg/column_mapping_spec.rb,
|
2246
|
+
spec/pg/result_spec.rb:
|
2247
|
+
Use instances of PG::ColumnMapping::CConverter as representation of
|
2248
|
+
the buildin converters instead of plain symbols. Symbol arguments
|
2249
|
+
are converted to these instances, for convenience.
|
2250
|
+
|
2251
|
+
This allows to store and retrieve additional informations to the
|
2252
|
+
conversion and also which conversions are supported.
|
2253
|
+
[ccf7ec4cd083]
|
2254
|
+
|
2255
|
+
* ext/pg_column_mapping.c:
|
2256
|
+
Don't depend on TYPE(obj)==T_DATA of a proc object for compat with
|
2257
|
+
rubinius.
|
2258
|
+
[f4f6f25434dd]
|
2259
|
+
|
2260
|
+
2013-06-29 Lars Kanis <lars@greiz-reinsdorf.de>
|
2261
|
+
|
2262
|
+
* ext/pg.c, ext/pg.h, ext/pg_column_mapping.c, ext/pg_result.c,
|
2263
|
+
spec/pg/result_spec.rb:
|
2264
|
+
Add column based type mapping.
|
2265
|
+
|
2266
|
+
This allowes to speed up result retrieval by directly converting
|
2267
|
+
integer and float values to proper ruby types.
|
2268
|
+
[71ddc0446d26]
|
2269
|
+
|
2270
|
+
2014-10-11 Michael Granger <ged@FaerieMUD.org>
|
2271
|
+
|
2272
|
+
* spec/helpers.rb, spec/pg/connection_spec.rb:
|
2273
|
+
Merged with 4cc778c5ead7
|
2274
|
+
[67bb0f34ca05]
|
2275
|
+
|
2276
|
+
2014-10-09 Lars Kanis <lars@greiz-reinsdorf.de>
|
2277
|
+
|
2278
|
+
* Rakefile, Rakefile.cross, ext/extconf.rb:
|
2279
|
+
gcc option -static-libgcc is required for i386-mingw32 build.
|
2280
|
+
|
2281
|
+
Don't use native pg_config for cross build. The paths were
|
2282
|
+
overwritten anyways until now.
|
2283
|
+
|
2284
|
+
Remove code that was only needed for static build to libpq.
|
2285
|
+
[4cc778c5ead7]
|
2286
|
+
|
2287
|
+
2014-10-09 Lars Kanis <kanis@comcard.de>
|
2288
|
+
|
2289
|
+
* Rakefile, Rakefile.cross, lib/pg.rb:
|
2290
|
+
Bundle libpq.dll into gem, because PostgreSQL-9.3 does no longer
|
2291
|
+
support static linking.
|
2292
|
+
[a37e27d356e5]
|
2293
|
+
|
2294
|
+
2014-10-04 Lars Kanis <lars@greiz-reinsdorf.de>
|
2295
|
+
|
2296
|
+
* spec/helpers.rb, spec/pg/connection_spec.rb:
|
2297
|
+
Add missing spec for hash form parameters to #exec_params .
|
2298
|
+
[04a6a0136a12]
|
2299
|
+
|
2300
|
+
2014-08-24 Lars Kanis <lars@greiz-reinsdorf.de>
|
2301
|
+
|
2302
|
+
* ext/pg_connection.c:
|
2303
|
+
Change all remaining Oid <-> VALUE conversions to UINT.
|
2304
|
+
[bae3928a7b09]
|
2305
|
+
|
2306
|
+
* .travis.yml:
|
2307
|
+
'gem install rake-compiler' is still needed for cross build.
|
2308
|
+
[591c4b0f2d69]
|
2309
|
+
|
2310
|
+
* lib/pg/connection.rb, spec/pg/connection_spec.rb:
|
2311
|
+
Add Connection.conndefaults_hash and add specs for conndefault
|
2312
|
+
variants.
|
2313
|
+
[2221655210da]
|
2314
|
+
|
2315
|
+
* .travis.yml:
|
2316
|
+
Travis uses the Gemfile, so 'gem install' is obsolete.
|
2317
|
+
[4feb9bc05f2f]
|
2318
|
+
|
2319
|
+
2014-08-23 Lars Kanis <lars@greiz-reinsdorf.de>
|
2320
|
+
|
2321
|
+
* .travis.yml, Rakefile.cross:
|
2322
|
+
Fix cross build for mingw32.
|
2323
|
+
[f97f4b2cf0c9]
|
2324
|
+
|
2325
|
+
* Rakefile.cross:
|
2326
|
+
Update postgresql and openssl versions for windows cross build.
|
2327
|
+
[f34546022095]
|
2328
|
+
|
2329
|
+
* .travis.yml:
|
2330
|
+
Fix travis build
|
2331
|
+
[daf56926b7d6]
|
2332
|
+
|
2333
|
+
* .travis.yml:
|
2334
|
+
Update travis test environment
|
2335
|
+
[cdb7c193ba29]
|
2336
|
+
|
2337
|
+
2014-08-22 Lars Kanis <lars@greiz-reinsdorf.de>
|
2338
|
+
|
2339
|
+
* Gemfile:
|
2340
|
+
Update Gemfile
|
2341
|
+
[a044c5419ff7]
|
2342
|
+
|
2343
|
+
* ext/extconf.rb, ext/pg_connection.c, lib/pg/connection.rb,
|
2344
|
+
spec/pg/connection_spec.rb:
|
2345
|
+
Test for PQconninfo availability and exclude conninfo and
|
2346
|
+
conninfo_hash otherwise.
|
2347
|
+
[f2692d108f46]
|
2348
|
+
|
2349
|
+
2014-08-20 Michael Granger <ged@FaerieMUD.org>
|
2350
|
+
|
2351
|
+
* spec/helpers.rb, spec/pg/connection_spec.rb:
|
2352
|
+
Check connection status with a matcher in specs
|
2353
|
+
[b32840b98e4b]
|
2354
|
+
|
2355
|
+
* ext/pg_connection.c, lib/pg/connection.rb,
|
2356
|
+
spec/pg/connection_spec.rb:
|
2357
|
+
Implement PG::Connection#conninfo and use it to test #188
|
2358
|
+
[affbd590e74e]
|
2359
|
+
|
2360
|
+
2014-07-23 Mina Naguib <mina.github@naguib.ca>
|
2361
|
+
|
2362
|
+
* ext/pg_result.c:
|
2363
|
+
Fix some type mismatches for Oid return values.
|
2364
|
+
|
2365
|
+
Thanks to Mina Naguib for the fix.
|
2366
|
+
[738d4c4125a1]
|
2367
|
+
|
2368
|
+
2014-07-23 Michael Granger <ged@FaerieMUD.org>
|
2369
|
+
|
2370
|
+
* Manifest.txt:
|
2371
|
+
Fix the path to spec helpers in the manifest
|
2372
|
+
[b3f9e94a258c]
|
2373
|
+
|
2374
|
+
2014-06-05 Michael Granger <ged@FaerieMUD.org>
|
2375
|
+
|
2376
|
+
* Merge with 6c2444dc63e1
|
2377
|
+
[d7160a9fb5dc]
|
2378
|
+
|
2379
|
+
2013-12-30 Lars Kanis <lars@greiz-reinsdorf.de>
|
2380
|
+
|
2381
|
+
* ext/pg_connection.c:
|
2382
|
+
Fix typo in documentation.
|
2383
|
+
[6c2444dc63e1]
|
2384
|
+
|
2385
|
+
2014-06-05 Michael Granger <ged@FaerieMUD.org>
|
2386
|
+
|
2387
|
+
* .rvm.gems, Rakefile, spec/helpers.rb, spec/lib/helpers.rb,
|
2388
|
+
spec/pg/connection_spec.rb, spec/pg/result_spec.rb, spec/pg_spec.rb:
|
2389
|
+
Convert specs to expect syntax for RSpec 3
|
2390
|
+
[c9108c846ab2]
|
2391
|
+
|
2392
|
+
2013-12-18 Michael Granger <ged@FaerieMUD.org>
|
2393
|
+
|
2394
|
+
* .hgtags:
|
2395
|
+
Added tag v0.17.1 for changeset 78846e47d87b
|
2396
|
+
[a5ab7f06aa1d]
|
2397
|
+
|
2398
|
+
* .hgsigs:
|
2399
|
+
Added signature for changeset 22d57e3a2b37
|
2400
|
+
[78846e47d87b] [v0.17.1]
|
2401
|
+
|
2402
|
+
* History.rdoc, lib/pg.rb:
|
2403
|
+
Bump the patch version, set the date in History.
|
2404
|
+
[22d57e3a2b37]
|
2405
|
+
|
1
2406
|
2013-12-17 Lars Kanis <kanis@comcard.de>
|
2
2407
|
|
3
2408
|
* History.rdoc:
|
4
2409
|
Update History file
|
5
|
-
[ebd0e140f78d]
|
2410
|
+
[ebd0e140f78d]
|
6
2411
|
|
7
2412
|
2013-12-16 larskanis <lars@greiz-reinsdorf.de>
|
8
2413
|
|
@@ -49,7 +2454,7 @@
|
|
49
2454
|
|
50
2455
|
* ext/pg_connection.c, lib/pg/connection.rb:
|
51
2456
|
Fix documentation for PG::Connection::conndefaults.
|
52
|
-
[9812218e0654]
|
2457
|
+
[9812218e0654]
|
53
2458
|
|
54
2459
|
2013-09-20 Lars Kanis <lars@greiz-reinsdorf.de>
|
55
2460
|
|