Camelia, the Perl 6 bug

IRC log for #cdk, 2010-05-14

| Channels | #cdk index | Today | | Search | Google Search | Plain-Text | summary

All times shown according to UTC.

Time Nick Message
05:15 sneumann joined #cdk
05:51 egonw joined #cdk
07:58 s_wolf joined #cdk
08:11 s_wolf joined #cdk
08:38 sneumann joined #cdk
09:13 maclean joined #cdk
09:14 maclean hello
09:14 zarah ni hao maclean
09:14 maclean $generic_greeting, robotic servant.
09:20 egonw hej maclean!
09:20 maclean hej
09:20 egonw good to see you on IRC :)
09:20 maclean :)
09:20 * egonw is working on hooking in test-signature into the build system
09:21 maclean I have a feeling that I should remove these .pack and .idx files before trying to pull again
09:22 egonw git clean ?
09:24 maclean nah, I just removed the files.
09:24 egonw sounds good to me :)
09:24 maclean using the git:// seems to work.
09:25 egonw good :)
09:25 egonw I have no clue if the http route is supposed to work or not...
09:25 egonw I guess it should
09:25 maclean Aargh! too well. I've aquired all your branches.
09:25 egonw as they have been braggin about being able to push over http too
09:25 egonw what git version do you have?
09:25 maclean uhm.
09:25 egonw maclean: that's fine
09:25 egonw they are not so many
09:26 egonw and there is no duplication anyway
09:26 * egonw is running: signatures-1.0-SNAPSHOT.jar
09:26 egonw nope, not that...
09:27 egonw what's up with my clipboard today?!?!
09:27 * egonw is running: ant -Dmodule=test-signature compile-module
09:27 maclean I didn't add any module info to the build.xml
09:27 maclean so that might not work.
09:28 egonw that's what I'm fixing :)
09:28 egonw no worries about that...
09:28 egonw this already works:
09:28 egonw ant -Dmodule=signature compile-module
09:28 maclean ah. good.
09:30 egonw ok, build works...
09:30 egonw now running it...
09:30 egonw ant -Dmodule=signature test-module
09:30 egonw Tests run: 48, Failures: 1, Errors: 26, Time elapsed: 1.207 sec
09:31 egonw java.lang.NoClassDefFoundError: org/openscience/cdk/tools/CDKHydrogenAdder
09:31 egonw ok, one more missing dep :)
09:31 maclean Yes, the tests, as usual will have a lot of deps.
09:31 egonw indeed
09:31 egonw Tests run: 48, Failures: 1, Errors: 7, Time elapsed: 2.738 sec
09:33 egonw Tests run: 48, Failures: 1, Errors: 0, Time elapsed: 31.75 sec
09:33 egonw one fail, is about unit test coverage
09:33 maclean Ah. Well, that's not unexpected.
09:34 egonw OK, lunch time here...
09:34 egonw pushed another commit, the one with the test-signature deps
09:34 egonw will be back later
09:34 maclean Ok.
10:41 egonw maclean: one more patch available... PMD testing
10:58 egonw joined #cdk
10:59 egonw network split :(
11:03 maclean ok, will just commit some javadoc improvements first
11:04 egonw ok, one more commit
11:04 egonw for javadoc.xml
11:05 maclean alright. push-pulled.
11:06 maclean I still have to write some descriptive text in the class javadoc for MoleculeSignature and AtomSignature.
11:11 maclean hmmm.
11:11 maclean BUILD FAILED
11:12 maclean Includesfile /Users/maclean/development/projec​ts/cdk/build/signature.javafiles not found.
11:15 egonw joined #cdk
11:15 egonw ok, and back again
11:16 egonw you know how to use @cdk.cite ?
11:17 maclean something like @cdk.cite TOR09 and then TOR09 is put somewhere else with the full paper ref. I did know, but I will look it up.
11:17 egonw doc/refs/cheminf.xml
11:18 egonw TOR09 ?
11:18 egonw is your signature work already published?
11:18 egonw but also please add @cdk.cite for AtomSignature and MoleculeSignature
11:18 maclean Hah. No, that was just an example :)
11:18 egonw pointing the the paper from ... ah :)
11:18 maclean My publication record is terrible.
11:19 egonw yeah, the most important thing
11:19 maclean I know, I know :(
11:33 egonw maclean: so, why do your signatures use [C] instead of C
11:33 egonw e.g. Fig 1 from the signature I paper
11:33 maclean The format was changed in subsequent papers.
11:33 egonw h=0 signature is 'C' in the paper, not '[C]' as with your impl
11:34 egonw oh, lovely...
11:34 maclean We've been trying to make the output as similar to Faulon's code as possible.
11:34 egonw well, ok, that's something to mention in the JavaDoc... :)
11:34 maclean I did try and make the code flexible enough to change format, but parsing was a nightmare.
11:34 egonw yes
11:34 egonw is a formal grammar possible for signatures?
11:35 maclean Ideally, something like cinfony should be able to do "CDK.signature(mol) == OBabel.signature(mol)"
11:35 egonw yeah, indeed
11:35 maclean yes, I think that is possible - what BNF?
11:36 egonw I have no one in mind
11:36 egonw one compatible with JavaCC perhaps :)
11:36 egonw
11:36 egonw junit.framework.AssertionFailedError: expected:<[C]> but was:<[C]>
11:36 egonw at junit.framework.Assert.fail(Assert.java:47)
11:36 egonw at junit.framework.Assert.fail​NotEquals(Assert.java:277)
11:36 egonw at junit.framework.Assert.ass​ertEquals(Assert.java:64)
11:36 egonw at junit.framework.Assert.ass​ertEquals(Assert.java:71)
11:36 egonw at org.openscience.cdk.signature.Atom​SignatureTest.testFigure1Signature​1Paper(AtomSignatureTest.java:150)
11:36 egonw at sun.reflect.NativeMethodAc​cessorImpl.invoke0(Native Method)
11:36 egonw at sun.reflect.NativeMethodAccessorImpl.in​voke(NativeMethodAccessorImpl.java:57)
11:36 egonw at sun.reflect.DelegatingMethodAccessorImpl.in​voke(DelegatingMethodAccessorImpl.java:43)
11:36 egonw at java.lang.reflect.Method.invoke(Method.java:616)
11:36 egonw at org.junit.runners.model.FrameworkMethod$1.​runReflectiveCall(FrameworkMethod.java:44)
11:36 egonw at org.junit.internal.runners.model.Reflectiv​eCallable.run(ReflectiveCallable.java:15)
11:36 egonw at org.junit.runners.model.FrameworkMethod.i​nvokeExplosively(FrameworkMethod.java:41)
11:36 egonw at org.junit.internal.runners.statements.Inv​okeMethod.evaluate(InvokeMethod.java:20)
11:37 egonw at org.junit.internal.runners.statements.R​unBefores.evaluate(RunBefores.java:28)
11:37 egonw at org.junit.internal.runners.statements.​RunAfters.evaluate(RunAfters.java:31)
11:37 egonw at org.junit.runners.BlockJUnit4ClassRunner.​runChild(BlockJUnit4ClassRunner.java:73)
11:37 egonw at org.junit.runners.BlockJUnit4ClassRunner.​runChild(BlockJUnit4ClassRunner.java:46)
11:37 egonw at org.junit.runners.ParentRunner.ru​nChildren(ParentRunner.java:180)
11:37 egonw at org.junit.runners.ParentRunner.a​ccess$000(ParentRunner.java:41)
11:37 egonw at org.junit.runners.ParentRunner$1​.evaluate(ParentRunner.java:173)
11:37 egonw at org.junit.internal.runners.statements.R​unBefores.evaluate(RunBefores.java:28)
11:37 egonw at org.junit.internal.runners.statements.​RunAfters.evaluate(RunAfters.java:31)
11:37 egonw at org.junit.runners.ParentRunne​r.run(ParentRunner.java:220)
11:37 egonw at org.eclipse.jdt.internal.junit4.runner.JUnit4T​estReference.run(JUnit4TestReference.java:46)
11:37 egonw at org.eclipse.jdt.internal.junit.runner.T​estExecution.run(TestExecution.java:38)
11:37 egonw at org.eclipse.jdt.internal.junit.runner.RemoteT​estRunner.runTests(RemoteTestRunner.java:467)
11:37 egonw at org.eclipse.jdt.internal.junit.runner.RemoteT​estRunner.runTests(RemoteTestRunner.java:683)
11:37 egonw at org.eclipse.jdt.internal.junit.runner.Remot​eTestRunner.run(RemoteTestRunner.java:390)
11:37 egonw at org.eclipse.jdt.internal.junit.runner.Remot​eTestRunner.main(RemoteTestRunner.java:197)
11:37 egonw oops
11:37 egonw sorry
11:37 egonw just wanted this bit:
11:37 egonw junit.framework.AssertionFailedError: expected:<[C]> but was:<[C]>
11:37 egonw lovely too...
11:37 egonw do you see the difference?
11:37 egonw oh, wait... hahahahah
11:37 egonw Assert.assertEquals("[C]", new AtomSignature(0, 0, methylnapthalene));
11:37 egonw sure
11:37 egonw upon making the error message it runs toString()...
11:37 egonw ok, that settles it...
11:38 egonw going to get some sleep
11:38 maclean :) good idea.
11:39 maclean Thanks for the reviews, egon.
11:41 egonw besides the [C] versus [C]
11:41 egonw did other things change too?
11:41 egonw I'm spotting a reordering in the signature
11:41 egonw comparing the signature1 paper and the impl
11:41 maclean Yes, colors are now 0-indexed instead of 1-indexed.
11:41 maclean Ah, there is a minimal/maximal difference.
11:42 maclean That may not be fully sorted yet...
11:42 egonw ok, that would likely explain
11:43 maclean So, in a canonical = minimal scheme, [C] would be canonical, while in a maximal one [H] would be...
11:43 maclean Which has implications for chirality, of course :)
11:45 maclean I'll put some of this detail in the javadocs.
12:01 CIA-30 cdk: Egon Willighagen master * ra1e1f1b / build.xml : Added JavaDoc testing to the QA task; removed module-uptodate which checked if the module was already compiled causing the target to be skipped - http://bit.ly/d7P4LM
12:03 egonw ok, update master or pull that last patch from my repos
12:03 egonw then try:
12:03 egonw ant -Dmodule=signature qa-module
12:05 egonw that will also run the PMD tests
12:05 egonw as well as the unit tests, and the JavaDoc
12:05 egonw .. checking
12:05 * egonw will be away for a while
16:34 s9asad joined #cdk
17:30 s9asad joined #cdk
17:37 s9asad left #cdk
17:44 s9asad joined #cdk
18:11 s9asad joined #cdk
18:20 s9asad joined #cdk
18:34 s9asad joined #cdk
19:10 carsten joined #cdk
19:28 egonw joined #cdk
22:51 sneumann joined #cdk
22:58 sneumann joined #cdk

| Channels | #cdk index | Today | | Search | Google Search | Plain-Text | summary