diff --git a/documentation/changelog.md b/documentation/changelog.md index a7c91f7..9193087 100644 --- a/documentation/changelog.md +++ b/documentation/changelog.md @@ -7,6 +7,7 @@ nav: ../ # History of Changes * [Introduction and explanation of symbols](#introduction) +* [Version 9.3-1104 (XXX)](#version_9.3-1104) * [Version 9.3-1100 (2013-11-01)](#version_9.3-1100) * [Version 9.2-1004 (2013-10-31)](#version_9.2-1004) * [Version 9.2-1003 (2013-07-08)](#version_9.2-1003) @@ -41,6 +42,599 @@ denote the various action types:![add](../media/img/add.jpg)=add, remove=remove, update=update *** + +## Version 9.3-1104 (XXX) + +commit 4881d971507b944c4b10018ddae9ab671081460f +Merge: 09eb202 9fdb65f +Author: Dave Cramer +Date: Fri Feb 14 08:17:15 2014 -0500 + + Merge pull request #120 from davecramer/REL9_3_STABLE + + Release version 9_3_STABLE 1101 + +commit 9fdb65f1851b6357d57c999485ae50b8bdc1f0f0 +Author: Dave Cramer +Date: Fri Feb 14 13:14:31 2014 +0000 + + incremented build to 1101 + +commit cc5dd475d9f61bf1b0cfb85f7cf9e2bde47e4569 +Author: Dave Cramer +Date: Thu Feb 13 12:04:22 2014 +0000 + + Do not merge the bump to 9.4 + This reverts commit aaa05872944d0a801530dc0f53fb6df84978d7f9. + +commit 87f9cde19e659eb8c44e1f8b2752434f90f4f9b4 +Merge: f46e9d1 8930805 +Author: Dave Cramer +Date: Tue Feb 11 10:36:15 2014 -0500 + + Merge pull request #113 from whitingjr/remove-toLowerCase-calls + + Added feature to disable column name sanitiser with a new property. + +commit 89308054f12ff064853d6fa126d693ca10a8aa84 +Merge: ced665e e3b821f +Author: Jeremy Whiting +Date: Mon Feb 3 15:33:58 2014 +0000 + + Merge branch 'remove-toLowerCase-calls' of github.com:whitingjr/pgjdbc into remove-toLowerCase-calls + +commit ced665ef19de63d71beca256bbe48b233221c97e +Author: Jeremy Whiting +Date: Wed Jan 22 19:06:31 2014 +0000 + + Added feature to disable column name sanitiser with a new property. Added test cases for existing default behaviour along with tests for behaviour when sanitiser is disabled. Can be configured via Driver or DataSource configuration property. + +commit 23fe2bba11f71cc5c459cb19b9d2679b2f17f599 +Author: Dave Cramer +Date: Tue Jan 28 09:14:41 2014 -0500 + + fixed compile error for 1.5 + +commit 6f1ee8994d3bb093add3235bb660782bee04e775 +Author: Craig Ringer +Date: Sat Jan 11 12:06:22 2014 +0800 + + Add a MainClass that tells the user they can't just run the JDBC driver + + After one too many reports of + "Failed to load Main-Class manifest attribute from postgresql-xxx.jar" + I'm submitting a dummy main-class that tells the user what they should + do instead. + + The message looks like: + + ------------ + PostgreSQL x.y JDBC4.1 (build bbbb) + Found in: jar:file:/path/to/postgresql-x.y-bbbb.jdbc41.jar!/org/postgresql/Driver.class + + The PgJDBC driver is not an executable Java program. + + You must install it according to the JDBC driver installation instructions for your application / container / appserver, then use it by specifying a JDBC URL of the form + jdbc:postgresql:// + or using an application specific method. + + See the PgJDBC documentation: http://jdbc.postgresql.org/documentation/head/index.html + + This command has had no effect. + ------------ + +commit 7ca8b3406fc1e9836627a3e534a93caad4a35670 +Author: Dave Cramer +Date: Fri Dec 20 07:23:42 2013 -0500 + + fixed bug PreparedStatement.getMetaData failed if result set was closed + reported by Emmanuel Guiton + +commit cb5f6ff9c0db1b8988bb7191f6e5c22856b5fa01 +Author: cchantep +Date: Thu Dec 12 15:54:55 2013 +0100 + + Base table more usefull than "" as basic table name + +commit df49f980cf2165d50e09930603c7434ea69578d4 +Author: Dave Cramer +Date: Mon Dec 9 15:22:58 2013 +0000 + + fixed driver fails to find foreign tables fix from plalg@hotmail.com + +commit fa078d6390d8b1c95b63cb65a801272c6bce9244 +Author: Heikki Linnakangas +Date: Thu Dec 5 20:21:50 2013 +0200 + + Fix various setQueryTimeout bugs. + + 1. If you call setQueryTimeout(5), wait 10 seconds, and call execute(), the + cancel timer has already expired, and the statement will be allowed to run + forever. Likewise, if you call setQueryTimeout(5), wait 4 seconds, and call + execute(), the statement will be canceled after only 1 second. + + 2. If you call setQueryTimeout on a PreparedStatement, and execute the same + statement several times, the timeout only takes affect on the first + statement. + + 3. If you call setQueryTimeout on one Statement, but don't execute it, the + timer will still fire, possible on an unrelated victim Statement. + + The root cause of all of these bugs was that the timer was started at the + setQueryTimeout() call, not on the execute() call. + + Also, remove the finally-block from the cancellation task's run-method, + because that might erroneously cancel the timer for the next query, if a + new query is started using the same statement fast enough. + +commit a146e24ea333799763ea6daebb158e80331428c4 +Author: Heikki Linnakangas +Date: Wed Nov 13 22:41:37 2013 +0200 + + Use StringBuffer to construct a string. + + This piece of code isn't performance-critical at all, but silences a + Coverity complaint. + +commit 3654d7a02917f51ed432fae6cf7e8f83f44465da +Author: Heikki Linnakangas +Date: Wed Nov 13 22:38:57 2013 +0200 + + Avoid integer overflow. + + The function returns long, but does the calculation first in int. If someone + sets the timeout to 600 hours in the URL, it will overflow, even though the + return value of the function is long and hence could return a larger value. + + To silence a Coverity complaint. + +commit cc733376dc16c3650e9cbb6faa130a3b5b0d1211 +Author: Heikki Linnakangas +Date: Wed Nov 13 22:31:43 2013 +0200 + + Plug some Statement leaks in metadata queries. + + These are fairly harmless, nobody calls these metadata methods frequently + enough for the leaks to matter, and a regular Statement doesn't hold onto + any server resources anyway. But let's appease Coverity. + +commit 095598c913d36bd04b0e585f95236b3f7447c2b8 +Author: Heikki Linnakangas +Date: Wed Nov 13 22:20:26 2013 +0200 + + Fix a couple of trivial Statement leaks. + + Found by Coverity. + +commit 1a090452e69c96f578babffe0e1f76214e5d5961 +Author: Heikki Linnakangas +Date: Wed Nov 13 21:32:54 2013 +0200 + + Make sure file is closed on exception. + + The system will eventually close the file anyway, and this read is highly + unlikely to throw an IOException in practice. + + Also, use RandomAccessFile.readFully(byte[]) to slurp the file into byte + array, rather than FileInputStream.read(byte[]). The latter would need to + be called in a loop to protect from short reads. + + Both issues were complained of by Coverity. + +commit 9605800e6aeae0637528171e92f8a3a3745f8c8f +Author: Heikki Linnakangas +Date: Wed Nov 13 21:23:51 2013 +0200 + + Mark a nested class as static. + + Coverity thinks that's cleaner. Also fix a comment typo while we're at it. + +commit 6f0d85a58da55f50fcffc3d984fd0e97be7172f9 +Author: Heikki Linnakangas +Date: Wed Nov 13 21:20:48 2013 +0200 + + Don't ignore the result of Statement.executeQuery(). + + Coverity thinks it's bad form. + +commit 72fd3b4b6e1d494ec392873de9c3eaab81337fd6 +Author: Stephen Nelson +Date: Tue Nov 12 21:29:01 2013 +0000 + + Generate a non-Maven JAR filename using build-time version and JDBC API level. + +commit 5a74ef1963007fc517c97bdc0c7abbaee1e8fa3a +Author: Dave Cramer +Date: Fri Nov 1 13:33:25 2013 +0000 + + opened HEAD for 9.4 development + +commit 3e3f860d0fbb562f2ae0393de5c3128f67f1361a +Author: Dave Cramer +Date: Thu Oct 31 13:29:18 2013 +0000 + + generate JDBC 4.1 string for 1.7 and 1.8 builds and set version to 1100 + +commit 65f6fef29c893fb8baf169d0a667834a8cc0d0c0 +Author: Stephen Nelson +Date: Wed Oct 30 20:56:42 2013 +0000 + + Generate a JDBC 4.1 version string for 1.7 and 1.8 build JDKs. + +commit f219944ebfaa4db4328f1391daa8bc02f884f779 +Author: Dave Cramer +Date: Wed Oct 30 11:09:31 2013 +0000 + + redo occurred spelling mistake + +commit 0923d2ba24c756e5a596f9dd1fe310fbf3c0fe83 +Author: Stephen Nelson +Date: Sat Oct 26 16:05:27 2013 +0100 + + Use 1.4 version of gpg plugin for signing Maven upload. + +commit f28b63662c59770f0ed14b669f7b2b5c6bc2c562 +Author: Stephen Nelson +Date: Sat Oct 26 15:00:27 2013 +0100 + + Revert "build for maven" + + This reverts commit e0a0baae0b4df7cb76c626e4479001fc973a454e. Resolved the conflicts on build.properties and build.xml to re-apply changes. + +commit 659c1c0181da467e0ec5bea0381be42ed304743d +Author: Stephen Nelson +Date: Sat Apr 13 00:20:36 2013 +0100 + + Build script changes to allow packaging and deployment to Maven central using maven-ant-tasks + + Updated build.properties to contain the sonatype urls. Updated build.xml so that gpg signing works for each accompanying artifact type. Updated pom.xml to allow templated group and artifact ids. + +commit f3933ccf4793d2f4597960aaed063cc16c79e862 +Author: Craig Ringer +Date: Fri Oct 25 16:02:37 2013 +0800 + + NonValidatingFactory should be included in both JDBC3 and JDBC4 + +commit b36b533ba4bcd8c785d8a0c9042e0bdd15af7d97 +Author: Michael McCaskill +Date: Wed Oct 23 15:07:24 2013 -0400 + + Use proper System property + + Using 'path.separator' results in malformed paths such as: + /default/dir:./postgresql/root.crt + This corrects the problem. + +commit 4a110edfc29b8c9854a77e431e43223084a28c1b +Author: Dave Cramer +Date: Thu Oct 17 09:10:15 2013 -0400 + + added sponsors + +commit 819d85fc65ee6eac984da54cea099e6525994ec9 +Author: Dave Cramer +Date: Thu Oct 17 08:29:07 2013 -0400 + + reset interrupted bit and throw unchecked exception if we get interrupted while trying to connect + +commit b0eed5b4bcb090cc4d689beb0a494b61fda747e4 +Author: Dave Cramer +Date: Tue Oct 15 06:54:23 2013 -0400 + + add functions to allow LargeObjectMaager to commit on close from Marc Cousin + +commit b72c4ae48cd84b53afbb1d57306e8ede4bda6ff5 +Author: Dave Cramer +Date: Thu Oct 10 17:54:14 2013 -0400 + + Fixed spelling of occured + +commit 902aa80c96538febae856a40e145a56e307d0641 +Author: Dave Cramer +Date: Thu Oct 10 17:33:23 2013 -0400 + + build for maven + +commit 46c1b50eb5d5fcfa76200454ae29662c3a7e85d1 +Author: tminglei +Date: Sat Aug 24 19:13:54 2013 +0800 + + add uuid array support (rename ArrayElementBuilder to ArrayAssistant) + +commit 719a1ecf3ae68f80265b494b6fe9f493457d5de4 +Author: tminglei +Date: Sat Aug 24 16:28:21 2013 +0800 + + add uuid array support + +commit e56d09d69245825c8ed30d4909f25cd3b68ee93b +Author: tminglei +Date: Sat Aug 24 14:14:31 2013 +0800 + + add uuid array support + +commit 9e5eec23643831e6a2a277f8e637c7321e553d64 +Author: Nick White +Date: Thu Apr 4 12:04:54 2013 +0100 + + allow the first query to be binary + +commit f46e9d113140c07a70c774e156ad041d02b182e2 +Merge: 6e69d27 de31165 +Author: Dave Cramer +Date: Thu Jan 30 03:15:34 2014 -0800 + + Merge pull request #116 from ignatov/patch-1 + + Better COLUMN_TYPE calculation + +commit de31165eb16e9716993ab883436570b72cf631c5 +Author: Sergey Ignatov +Date: Thu Jan 30 15:00:56 2014 +0400 + + test added + +commit 1548c0fb8213fef2464b448a56ee3fdcfa05b3a0 +Author: Sergey Ignatov +Date: Wed Jan 29 18:22:11 2014 +0300 + + Better COLUMN_TYPE calculation + +commit 6e69d2740a4f0c3fb4df408bae8b1885b842dd56 +Merge: 1ce28a3 27147b3 +Author: Dave Cramer +Date: Tue Jan 28 15:31:42 2014 -0800 + + Merge pull request #114 from davecramer/master + + fixed compile error for 1.5 + +commit 27147b35def4019f55076f4b139cb79c20a2e6a7 +Author: Dave Cramer +Date: Tue Jan 28 09:14:41 2014 -0500 + + fixed compile error for 1.5 + +commit e3b821fc20eed5658fdb771f2797adde8256d6f7 +Author: Jeremy Whiting +Date: Wed Jan 22 19:06:31 2014 +0000 + + Added feature to disable column name sanitiser with a new property. Added test cases for existing default behaviour along with tests for behaviour when sanitiser is disabled. Can be configured via Driver or DataSource configuration property. + +commit 1ce28a3a032dd22bcadbd9852a17537ac3e04de7 +Merge: bf7e866 9ef0411 +Author: Dave Cramer +Date: Sun Jan 12 03:46:03 2014 -0800 + + Merge pull request #112 from ringerc/AddMainClass + + Add a MainClass that tells the user they can't just run the JDBC driver + +commit 9ef04115559d1292b06e0a38f76742434ac25167 +Author: Craig Ringer +Date: Sat Jan 11 12:06:22 2014 +0800 + + Add a MainClass that tells the user they can't just run the JDBC driver + + After one too many reports of + "Failed to load Main-Class manifest attribute from postgresql-xxx.jar" + I'm submitting a dummy main-class that tells the user what they should + do instead. + + The message looks like: + + ------------ + PostgreSQL x.y JDBC4.1 (build bbbb) + Found in: jar:file:/path/to/postgresql-x.y-bbbb.jdbc41.jar!/org/postgresql/Driver.class + + The PgJDBC driver is not an executable Java program. + + You must install it according to the JDBC driver installation instructions for your application / container / appserver, then use it by specifying a JDBC URL of the form + jdbc:postgresql:// + or using an application specific method. + + See the PgJDBC documentation: http://jdbc.postgresql.org/documentation/head/index.html + + This command has had no effect. + ------------ + +commit bf7e866ed5f051f4d2d9548c2976daf80420c3b9 +Merge: cea966c 2bcb5bd +Author: Dave Cramer +Date: Thu Dec 26 13:59:00 2013 -0800 + + Merge pull request #107 from cchantep/rs-basic-tblname + + Basic table name for resultset metadata + +commit cea966cce2595ce0a37bb7dbb42dd94c88fdd534 +Merge: 1463e98 8bd3f87 +Author: Dave Cramer +Date: Fri Dec 20 07:09:05 2013 -0800 + + Merge pull request #109 from davecramer/REL9_3_STABLE + + fixed bug PreparedStatement.getMetaData failed if result set was closed + +commit 8bd3f87c88bfd8e22f2884dcddf34ddf60804130 +Author: Dave Cramer +Date: Fri Dec 20 10:07:15 2013 -0500 + + fixed bug PreparedStatement.getMetaData failed if result set was closed + reported by Emmanuel Guiton #bugfix# + +commit 1463e985e233123b16d84394423aab345f2ebf6f +Merge: c007fde 3c57c9c +Author: Dave Cramer +Date: Fri Dec 20 04:27:15 2013 -0800 + + Merge pull request #108 from davecramer/master + + fixed bug PreparedStatement.getMetaData failed if result set was closed + +commit 3c57c9ca171adbb664f3a94d4394086114004684 +Author: Dave Cramer +Date: Fri Dec 20 07:23:42 2013 -0500 + + fixed bug PreparedStatement.getMetaData failed if result set was closed + reported by Emmanuel Guiton + +commit 2bcb5bd17e872d94056d5f338468ef2fdaca36b0 +Author: cchantep +Date: Thu Dec 12 15:54:55 2013 +0100 + + Base table more usefull than "" as basic table name + +commit c007fdedcbc667dddcbb6d5b2732eaaf1d7ae957 +Merge: a3bdcfd 2c6eb23 +Author: Dave Cramer +Date: Mon Dec 9 08:15:40 2013 -0800 + + Merge pull request #105 from hlinnaka/querytimeout-fixes + + Fix various setQueryTimeout bugs. + +commit a3bdcfd87b7770f2f629fcd16538d8a4ccdce558 +Merge: a3fa0dd 9fffe2c +Author: Dave Cramer +Date: Mon Dec 9 08:15:18 2013 -0800 + + Merge pull request #106 from davecramer/master + + fixed driver fails to find foreign tables fix from plalg@hotmail.com + +commit 9fffe2c544de4860a6986105e861c9cfeb2ced9e +Author: Dave Cramer +Date: Mon Dec 9 15:22:58 2013 +0000 + + fixed driver fails to find foreign tables fix from plalg@hotmail.com + +commit 2c6eb23eeb8e5a77ef58d9e9840f5ee6078ec366 +Author: Heikki Linnakangas +Date: Thu Dec 5 20:21:50 2013 +0200 + + Fix various setQueryTimeout bugs. + + 1. If you call setQueryTimeout(5), wait 10 seconds, and call execute(), the + cancel timer has already expired, and the statement will be allowed to run + forever. Likewise, if you call setQueryTimeout(5), wait 4 seconds, and call + execute(), the statement will be canceled after only 1 second. + + 2. If you call setQueryTimeout on a PreparedStatement, and execute the same + statement several times, the timeout only takes affect on the first + statement. + + 3. If you call setQueryTimeout on one Statement, but don't execute it, the + timer will still fire, possible on an unrelated victim Statement. + + The root cause of all of these bugs was that the timer was started at the + setQueryTimeout() call, not on the execute() call. + + Also, remove the finally-block from the cancellation task's run-method, + because that might erroneously cancel the timer for the next query, if a + new query is started using the same statement fast enough. + +commit a3fa0dde11255d580129dd39c3d5322701b16563 +Merge: 6d25fc8 4baac5d +Author: Dave Cramer +Date: Fri Nov 15 06:16:09 2013 -0800 + + Merge pull request #102 from hlinnaka/coverity-fixes + + Coverity fixes + +commit 6d25fc897e240956318e692f6cebe2ceddcd2f67 +Merge: aaa0587 270d357 +Author: Dave Cramer +Date: Fri Nov 15 05:51:08 2013 -0800 + + Merge pull request #101 from lordnelson/jar-rename + + Create a version string when building the JAR + +commit 4baac5d95f927d20baac61fab03450f4cab6e051 +Author: Heikki Linnakangas +Date: Wed Nov 13 22:41:37 2013 +0200 + + Use StringBuffer to construct a string. + + This piece of code isn't performance-critical at all, but silences a + Coverity complaint. + +commit 5fa0ee94f261599bd5434b564460be95c38b3cab +Author: Heikki Linnakangas +Date: Wed Nov 13 22:38:57 2013 +0200 + + Avoid integer overflow. + + The function returns long, but does the calculation first in int. If someone + sets the timeout to 600 hours in the URL, it will overflow, even though the + return value of the function is long and hence could return a larger value. + + To silence a Coverity complaint. + +commit ec0c79a3b7af00e24e39600e15cd14c7745af959 +Author: Heikki Linnakangas +Date: Wed Nov 13 22:31:43 2013 +0200 + + Plug some Statement leaks in metadata queries. + + These are fairly harmless, nobody calls these metadata methods frequently + enough for the leaks to matter, and a regular Statement doesn't hold onto + any server resources anyway. But let's appease Coverity. + +commit 0306ae033b7005f266333f41b2ee998af0ec2309 +Author: Heikki Linnakangas +Date: Wed Nov 13 22:20:26 2013 +0200 + + Fix a couple of trivial Statement leaks. + + Found by Coverity. + +commit 9aeee3a6c5b91b76565ce99161c5efcd13b9696b +Author: Heikki Linnakangas +Date: Wed Nov 13 21:32:54 2013 +0200 + + Make sure file is closed on exception. + + The system will eventually close the file anyway, and this read is highly + unlikely to throw an IOException in practice. + + Also, use RandomAccessFile.readFully(byte[]) to slurp the file into byte + array, rather than FileInputStream.read(byte[]). The latter would need to + be called in a loop to protect from short reads. + + Both issues were complained of by Coverity. + +commit 26e5fde35ca5052025ba847649ff0f763da5dfea +Author: Heikki Linnakangas +Date: Wed Nov 13 21:23:51 2013 +0200 + + Mark a nested class as static. + + Coverity thinks that's cleaner. Also fix a comment typo while we're at it. + +commit eb8d050a549176336530a5a5c4c2aeb097e14529 +Author: Heikki Linnakangas +Date: Wed Nov 13 21:20:48 2013 +0200 + + Don't ignore the result of Statement.executeQuery(). + + Coverity thinks it's bad form. + +commit 270d35711746b718a2e476af0caa373a94f47a05 +Author: Stephen Nelson +Date: Tue Nov 12 21:29:01 2013 +0000 + + Generate a non-Maven JAR filename using build-time version and JDBC API level. + +commit aaa05872944d0a801530dc0f53fb6df84978d7f9 +Author: Dave Cramer +Date: Fri Nov 1 13:33:25 2013 +0000 + + opened HEAD for 9.4 development + + ## Version 9.3-1100 (2013-11-01)