Camelia, the Perl 6 bug

IRC log for #cdk, 2008-05-09

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

All times shown according to UTC.

Time Nick Message
05:55 jbrefort joined #cdk
06:16 malick joined #cdk
06:59 carsten joined #cdk
07:49 CIA-31 bioclipse: ospjuth * r5604 /bioclipse2/trunk/plugins/net.bioclipse.cdk10.​jchempaint/src/net/bioclipse/cdk10/jchempaint/​ui/editor/JCPMultiPageEditorContributor.java:
07:49 CIA-31 bioclipse: Workaround for a ClassCastException that tried to add ICDKChangeListener to the
07:49 CIA-31 bioclipse: IJCPBasedEditor and not to the JCPPAge of it. Occurs if Bioclipse is started
07:49 CIA-31 bioclipse: with an IJCPBasedEditor already open. Placed FIXME in
07:49 CIA-31 bioclipse: JCPMultiPageEditorContributor.
07:55 jonalv joined #cdk
08:15 rojasm joined #cdk
08:58 masak joined #cdk
09:11 rojasm updating cdk the ReaderFactory doesn't find some import classes which I suppose are missing or they don't exist anymore
09:11 rojasm PubChemcompoundXMLFormat
09:11 rojasm and
09:11 rojasm PubChemSubstanceXMLFormat
09:13 CIA-31 cdk: archvile18 * r10930 /cdk/trunk/src/ (4 files in 4 dirs): Fix for bug [1959516] "Pyrrole nitrogen not typed PLANAR3".
10:00 Andreas1 joined #cdk
10:00 Andreas1 hi
10:32 CIA-31 bioclipse: jonalv * r5605 /bioclipse2/trunk/plugins/ (5 files in 4 dirs): Structuredb instances now survives restart of Bioclipse correctly and can not only be created but also deleted
10:55 CIA-31 bioclipse: ospjuth * r5606 /bioclipse2/trunk/plugins/net.bioc​lipse.cdk10.jchempaint/src/net/bio​clipse/cdk10/jchempaint/ui/editor/ (4 files): Updated for getting Coloring to work. Added rePaint() in JCPPage, placed TODO:verify for egonw in stateChanged(...).
11:44 CIA-31 bioclipse: ospjuth * r5607 /bioclipse2/trunk/plugins/net.bioclipse.cd​k10.jchempaint/src/net/bioclipse/cdk10/jch​empaint/ui/editor/MDLMolfileEditor.java: Added EDITOR_ID to class.
13:18 jpansanel moin
13:21 carsten joined #cdk
13:32 malick joined #cdk
13:40 CIA-31 bioclipse: carl_masak * r5608 /bioclipse2/trunk/plugins/net.bioclipse.cd​k.business/src/net/bioclipse/cdk/business/ (CDKManager.java ICDKManager.java):
13:40 CIA-31 bioclipse: [ICDKManager, CDKManager]
13:40 CIA-31 bioclipse: * name change: createMoleculeFromSMILES -> moleculeFromSmiles
13:40 CIA-31 bioclipse: * whitespace and line length
14:05 CIA-31 bioclipse: carl_masak * r5609 /bioclipse2/trunk/plugins/net.bioclips​e.cdk.business.test/src/net/bioclipse/​cdk/business/test/CDKManagerTest.java:
14:05 CIA-31 bioclipse: [CDKManagerTest]
14:05 CIA-31 bioclipse: * chased name change: createMoleculeFromSMILES -> moleculeFromSmiles
14:06 masak language question: how would you describe the relation between a molecule and a substructure of that molecule?
14:07 masak is it valid to say that the former 'contains' the latter?
14:13 Andreas1 "The substructure is present in the molecule." should work as well
14:13 Andreas1 but "contains" sounds good to me too
14:13 masak good.
14:14 masak it has the added advantage of being the same relation as between Strings and their substrings in Java
14:15 Andreas1 from a java perspective I'd definitely use ".contains()"
14:16 masak mm
14:16 Andreas1 one could think of ".match()" in terms of "the molecules matches the substructure query" as well, though
14:16 jonalv okey I'll go for that then :)
14:19 masak I'd do "match()" if wildcards are involved somehow
14:19 jonalv no wildcards
14:19 masak jonalv: .contains(), then
14:20 jonalv the call would look like this: structuredb.retrieveStructuresContaining( queryStructure );
14:21 jonalv and that will call cdk.contains( checkedMolecule, queryMolecule )
14:21 jonalv or something like that
14:21 jonalv what do you think?
14:21 jonalv the second one isn't great right?
14:24 Andreas1 .containsSubstructure() or .querySubstructure() or .searchSubstructure() ?
14:24 jonalv .containsSubstructure is probably much better
14:25 masak I like `.allStructuresContaining()` for the first one
14:25 masak 'retrieve' sounds boring
14:25 jonalv I call all my methods returning a list for retrieve something...
14:25 masak it's long, hard to spell and not very fun
14:26 masak you're probably using it because 'get' is taken :)
14:26 masak bad sign
14:26 jonalv who is gonna spell it when there is tab completion and yes get sounds like something else
14:26 masak don't count on tab completion to save you
14:26 masak besides, short names are sexy
14:27 masak point is, you don't need 'get' either
14:27 jonalv not if you don't understand what they do...
14:27 Andreas1 I prefer long, more telling names
14:27 masak just do 'all' like above
14:27 masak Andreas1: you and jonalv will get along, then :)
14:27 CIA-31 bioclipse: ospjuth * r5610 /bioclipse2/trunk/plugins/ (11 files in 6 dirs): Working on getting DC on entries in StructureTable open in JCP.
14:27 Andreas1 heh
14:28 jonalv masak: please have a look at the api for the strucutredb in bioclipse and see what you think when seeing it all at the same time
14:28 masak jonalv: will do
14:44 masak jonalv: ok, I'm ready to give some feedback. do you want to continue the discussion here or on #bioclipse?
14:44 jonalv hm maybe not here
14:44 jonalv what about IRL? :)
14:44 masak hm
14:55 malick left #cdk
15:00 CIA-31 bioclipse: jonalv * r5611 /bioclipse2/trunk/plugins/ (3 files in 2 dirs): added cdk.containsSubstructure
15:14 Andreas1 How would I look up bug #956929 in the sf bug tracker?
15:22 Andreas1 when I search the trackers on sf for the bug id it gives me zero results
15:23 Andreas1 but there is a test case which references this id
15:31 egonw joined #cdk
15:35 CIA-31 cdk: egonw * r10931 /cdk/trunk/src/test/org/openscienc​e/cdk/io/MDLV2000ReaderTest.java: Fixed test: order is not kept per definition
15:37 CIA-31 cdk: egonw * r10932 /cdk/trunk/src/test/org/openscienc​e/cdk/io/MDLV2000ReaderTest.java: Cleaned up code; assert additional assertion
15:38 Andreas1 Hi Egon, could you read my comment to bug to the latest bug report pls?
15:39 egonw saw that...
15:39 egonw was that a new patch?
15:39 egonw that separate from what you emailed yesterday?
15:39 Andreas1 yes, it's a new problem and a new patch for it
15:40 CIA-31 cdk: egonw * r10933 /cdk/trunk/src/main/org/openscie​nce/cdk/io/MDLV2000Reader.java: Fixed reading of isotope information: read correct columns
15:41 CIA-31 bioclipse: jonalv * r5612 /bioclipse2/trunk/plugins/ (3 files in 2 dirs): Renamed a few methods for structuredb. Hopes this makes things even better :)
15:41 CIA-31 cdk: egonw * r10934 /cdk/trunk/src/data/mdl/hisotopes.mol: Fixed mass number of T
15:42 egonw Andreas1: can you briefly explain this case:
15:42 egonw case....
15:42 egonw (no :)
15:43 egonw do I understand correctly that if a ring already contains a "N.planar3", the other N atoms are N.sp2?
15:43 egonw that requires the other ring atoms to be alreadt perceived, which is not generally true
15:44 CIA-31 cdk: egonw * r10935 /cdk/trunk/src/main/org/openscie​nce/cdk/io/MDLV2000Reader.java: Added IO option (default: true) to interpret T and D as hydrogen isotopes, IAOI 'M ISO' was used in conjunction to set correct mass numbers (closes #1826577)
15:44 Andreas1 sure, that's the case for imidazole [nH]1cncc1
15:44 egonw right... but let's consider this SMILES:
15:44 egonw n1cc[nH]c1
15:44 egonw then the containsAtomType(ring, "N.planar3") return false...
15:44 Andreas1 wait, I'll try that
15:45 CIA-31 cdk: egonw * r10936 /cdk/trunk/src/test/org/opensci​ence/cdk/io/MDLReaderTest.java: Removed test for MDL V2000 mol file (which was already present in MDLV2000ReaderTest anyway)
15:48 CIA-31 cdk: egonw * r10937 /cdk/trunk/src/main/org/openscie​nce/cdk/io/MDLV3000Reader.java: Fixed reader of the header of MDL V3000 files: did not read the 4th line yet
15:48 Andreas1 true, that smiles doesn't work
15:49 egonw no worries...
15:49 egonw we'll come up with something...
15:49 egonw the way I solved this kind of problem is to perceive the other ring atoms...
15:49 egonw that is...
15:49 egonw we need a helper method like isNPlanar3()
15:49 jbrefort joined #cdk
15:50 egonw which decides this on the fly...
15:50 Andreas1 actually, to make it 100% correct we need to change the SmilesParser
15:50 egonw yes, see one of my last bug reports...
15:50 masak egonw: I know I've asked this before, but to what extent can CDK convert a molecule to a IUPAC name for that molecule?
15:50 egonw we need to make the SmilesParser preserve hardcoded info like implicit hydrogens...
15:50 egonw Andreas1: I know how to do this, and will try to find some time to implement that
15:51 egonw masak: CDK can't...
15:51 masak ok
15:51 egonw it's a difficult problem
15:51 masak guess so
15:51 Andreas1 yes, but I just realized that n1cc[nH]c1 is actually percieved as if it would be [nH]1ccnc1...
15:51 jonalv masak: how far along are you? :)
15:51 masak jonalv: insignificantly so...
15:51 masak I can see the problems up ahead
15:52 egonw Andreas1: might you please add a second unit test using the alternative SMILES?
15:52 Andreas1 so, yes, the SmilesParser needs to set the implicit hydrogens...
15:52 egonw yes
15:52 Andreas1 and the atom types needs to use this info too...
15:52 egonw will see to that
15:52 egonw yes
15:52 Andreas1 *typer
15:52 egonw unit tests for that might be good to have...
15:52 egonw ok, will have to go offline now...
15:53 egonw and make dinner
15:53 egonw so, for now... patch declined...
15:53 Andreas1 k, I'll commit the tests for the alternative smiles
15:53 egonw ok, thanx
15:53 Andreas1 cu later
15:56 jonalv joined #cdk
16:09 CIA-31 bioclipse: jonalv * r5613 /bioclipse2/trunk/plugins/ (3 files in 2 dirs): Created test for substructure-search in structuredb
16:10 CIA-31 cdk: archvile18 * r10938 /cdk/trunk/src/test/org/openscience​/cdk/smiles/SmilesParserTest.java: Test cases for bug #1960990.
18:20 egonw joined #cdk
19:18 egonw_ joined #cdk
19:53 CIA-31 cdk: egonw * r10939 /cdk/trunk/src/test/org/openscience​/cdk/smiles/SmilesParserTest.java: Added test for preserving hard coded hydrogens
20:24 rojasm joined #cdk
20:31 CIA-31 cdk: egonw * r10940 /cdk/trunk/src/test/org/openscience​/cdk/smiles/SmilesParserTest.java: Fixed test. Was originally added in r8332, in apparently a black out :)
20:32 CIA-31 cdk: egonw * r10941 /cdk/trunk/src/test/org/openscience​/cdk/smiles/SmilesParserTest.java: Fixed atom index for second nitrogen
20:32 CIA-31 cdk: egonw * r10942 /cdk/trunk/src/main/org/openscien​ce/cdk/smiles/SmilesParser.java: Implemented preserving hardcoded implicit hydrogen counts (closes #1957127)
20:49 rojasm joined #cdk

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