Camelia, the Perl 6 bug

IRC log for #cdk, 2012-04-01

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

All times shown according to UTC.

Time Nick Message
04:41 egonw joined #cdk
05:27 CIA-121 cdk: Egon Willighagen master * r0040b00 / src/META-INF/test-standard.cdkdepends :
05:27 CIA-121 cdk: Added a new dependency
05:27 CIA-121 cdk: Change-Id: I2b82b734137a5f351cd5eed82c9124aac548327f - http://git.io/AE62lg
07:33 CIA-121 cdk: Rajarshi Guha cdk-1.4.x * r57a9c3f / src/test/org/openscience/cdk/finger​print/PubchemFingerprinterTest.java :
07:33 CIA-121 cdk: Added a unit test to check for multithreaded usage in PubchemFingerprint. Tests for bug 3510588
07:33 CIA-121 cdk: Change-Id: I2ee5c46b49846b1072949d9c0c06ec715fad8cb7
07:33 CIA-121 cdk: Signed-off-by: Egon Willighagen <egonw@users.sourceforge.net> - http://git.io/zjuA3w
07:33 CIA-121 cdk: Egon Willighagen cdk-1.4.x * r7c5d18b / src/test/org/openscience/cdk/finger​print/PubchemFingerprinterTest.java :
07:33 CIA-121 cdk: Removed to unused imports, one of which causes trouble with free Java systems
07:33 CIA-121 cdk: Change-Id: Ida389c59ad41b605b4243b557421c88c09cd02e4 - http://git.io/nqeX5Q
07:33 CIA-121 cdk: Rajarshi Guha cdk-1.4.x * r9bd6b45 / src/main/org/openscience/cdk/fing​erprint/PubchemFingerprinter.java :
07:33 CIA-121 cdk: Updated PubchemFingperinter to not use static variables. Allows it to be used in multithreaded scenarios
07:33 CIA-121 cdk: Change-Id: Iaa9481fd1fa5793cb5fe3b449219aeb7958a983d
07:33 CIA-121 cdk: Signed-off-by: Egon Willighagen <egonw@users.sourceforge.net> - http://git.io/7q2QnQ
07:37 slyrus egonw: the linefeeds in InChIGenerator.java are messed up
07:38 slyrus they're 0x0a 0x0d instead of 0x0d 0d0a or, perhaps better yet, 0x0a
07:46 egonw can you write a unit test for that?
07:46 egonw I guess these are \n and \r ?
07:47 egonw btw, you complain mostly about maven...
07:47 egonw in your post... not about the cdk
07:47 egonw that's fine to me... I'm not a maven fan either :)
07:48 slyrus heh
07:48 slyrus I can complain about anything!
07:49 egonw yes, please do :)
07:49 egonw no complaining, no progress
07:49 slyrus I haven't gotten started on my gripes about the SDG. But that hardly seems productive...
07:50 egonw we should try to convert that in plans
07:50 egonw student projects...
07:50 egonw clear description of a small bit of the problem
07:50 egonw like E/Z visualization
07:50 egonw and which steps must be taken to fix it
07:50 egonw something like:
07:50 slyrus E/Z and R/S (via wedged/hashed bonds) are two imporant steps
07:50 slyrus important even
07:53 slyrus I also plan on writing a blog post on my clojure-cdk experiences. the more languages the merrier!
07:53 egonw absolutely!
07:53 slyrus do we really need unit tests for bogus line feeds or can we just fix them as we notice them?
07:54 egonw not sure where the bogus line feeds are...
07:54 egonw detail?
07:54 egonw we don't want them to get reintroduced
07:54 egonw assuming it was in generated InChI output...
07:54 egonw but maybe you meant something different?
07:55 slyrus no, this is in the InChiGenerator.java source file
07:55 slyrus ICBW but it looks bogus to me
08:00 egonw which line?
08:00 slyrus all of them!
08:01 egonw oh... you mean just the EOLs in the source code?
08:02 slyrus right
08:31 egonw a commit just fixing new lines (but no whitespace or other stuff) is most welcome
08:45 CIA-121 cdk: Rajarshi Guha master * rd0e5535 / src/test/org/openscience/cdk/finger​print/PubchemFingerprinterTest.java :
08:45 CIA-121 cdk: Added a unit test to check for multithreaded usage in PubchemFingerprint. Tests for bug 3510588
08:45 CIA-121 cdk: Change-Id: I2ee5c46b49846b1072949d9c0c06ec715fad8cb7
08:45 CIA-121 cdk: Signed-off-by: Egon Willighagen <egonw@users.sourceforge.net> - http://git.io/RN7Fqg
08:45 CIA-121 cdk: Egon Willighagen master * re5b014b / src/test/org/openscience/cdk/finger​print/PubchemFingerprinterTest.java :
08:45 CIA-121 cdk: Removed to unused imports, one of which causes trouble with free Java systems
08:45 CIA-121 cdk: Change-Id: Ida389c59ad41b605b4243b557421c88c09cd02e4 - http://git.io/dIZQ_w
08:45 CIA-121 cdk: Rajarshi Guha master * r171c06b / src/main/org/openscience/cdk/fing​erprint/PubchemFingerprinter.java :
08:45 CIA-121 cdk: Updated PubchemFingperinter to not use static variables. Allows it to be used in multithreaded scenarios
08:46 CIA-121 cdk: Change-Id: Iaa9481fd1fa5793cb5fe3b449219aeb7958a983d
08:46 CIA-121 cdk: Signed-off-by: Egon Willighagen <egonw@users.sourceforge.net>
08:46 CIA-121 cdk: Conflicts:
08:46 CIA-121 cdk:  src/main/org/openscience/cdk/fing​erprint/PubchemFingerprinter.java - http://git.io/sQqeQg
08:46 CIA-121 cdk: Egon Willighagen master * r9fbf219 / src/main/org/openscience/cdk/fing​erprint/PubchemFingerprinter.java :
08:46 CIA-121 cdk: Removed the unneeded catching of CDKException
08:46 CIA-121 cdk: Change-Id: Ie571fefc7e93903c13c74914bda4593b35ac7208 - http://git.io/VTfftg
09:22 CIA-121 cdk: Egon Willighagen cdk-1.4.x * rdf12e2e / build.xml :
09:22 CIA-121 cdk: Added the missing cdk-test.jar (patch by Jonty Lawson)
09:22 CIA-121 cdk: Change-Id: I0408c9e03ace32aea9a7b9c4a7f031b4cc1dc62e - http://git.io/VPL6zQ
09:22 CIA-121 cdk: Egon Willighagen cdk-1.4.x * r059da00 / build.xml :
09:22 CIA-121 cdk: But the test data should be added too
09:22 CIA-121 cdk: Change-Id: I926c869d94562bb6b086b5556b02359c46abc597 - http://git.io/pNBs_g
10:34 egonw joined #cdk
10:42 CIA-121 cdk: Egon Willighagen master * r02fec4a / build.xml :
10:42 CIA-121 cdk: Added the missing cdk-test.jar (patch by Jonty Lawson)
10:42 CIA-121 cdk: Change-Id: I0408c9e03ace32aea9a7b9c4a7f031b4cc1dc62e - http://git.io/Mu8Aaw
10:42 CIA-121 cdk: Egon Willighagen master * rf2080e1 / build.xml :
10:42 CIA-121 cdk: But the test data should be added too
10:42 CIA-121 cdk: Change-Id: I926c869d94562bb6b086b5556b02359c46abc597 - http://git.io/SMlVhQ
13:28 egonw joined #cdk
14:33 slyrus egonw: good morning. is the preferred EOL LF (unix) or CRLF (DOS)?
14:34 egonw good question... normally SVN/Git handles those things
14:35 egonw what do they other files have?
14:36 slyrus LF
14:36 slyrus wasn't sure if there were some conventions I should know about
14:45 egonw ok, then go for LF too
14:47 slyrus egonw: not sure what the best way to get this into the queue is, but for the moment it's at: https://github.com/slyrus/cdk/commit/53​ff61418f445d84753a6b85150fe12c665f0760
14:48 egonw file a report in the patch tracker on SF, pointing to that commit?
14:49 egonw that would work for me...
14:49 egonw I need to finish some other stuff before I can do further CDK hacking today
14:49 slyrus OK. InChiGeneratorFactory has the same problem. I'll add another patch and put it in the tracker.
15:01 egonw ok
15:03 slyrus when you're not too busy, I'll have to get you to walk me through testing the inchi stuff. It seems to be loading OK, but I get a #<net.sf.jniinchi.INCHI_RET ERROR {4E561369}> instead of an inchi...
15:05 slyrus oh, never mind... "Accepted unusual valence(s): C(3); Cannot process aromatic bonds"
15:06 slyrus and we're one step farther down the rabbit hole... lisp, calling java, calling native C++
15:35 egonw yeah, that is somethng I have high on my agenda
15:35 egonw we are working hard towards an improved tool to place double bonds...
15:36 slyrus ah, we just call them aromatic ATM?
15:37 slyrus rather than picking a particular kekule form?
15:37 egonw depends on the input
15:37 egonw I am not sure what the inchi library does with 'aromatic' bonds...
15:37 egonw never tried...
15:37 egonw MDL molfiles themselves don't have the concept
15:37 egonw there is a concept in the matching *query* format
15:44 slyrus a simplified example of what I'm trying to do is: (get-inchi (parse-smiles-string "n1ccccc1"))
15:47 egonw understood
15:47 egonw deducing double bond placement should go in between
15:47 egonw I think the SMILES parser may do that in the future
15:51 slyrus yet benzene (as c1ccccc1) works
15:51 egonw can you past the atom types detected for the n1ccccc1 ?
15:52 egonw or just for the nitrogen?
15:54 slyrus sure. also, now I need the chemicl molecule <--> CDK molecule bridge so I can see if my SMILES parser is any better :)
15:54 egonw fair
15:55 egonw have to go, bbl
15:59 slyrus ok, thanks for the help!
15:59 slyrus (describe (#"getAtom" *pyridine* 0)) says:
15:59 egonw listening...
15:59 slyrus The wrapped Java object is an instance of org.openscience.cdk.Atom: "Atom(139354187, S:N, H:0, AtomType(139354187, N:N.sp2, FC:0, H:SP2, NC:2, EV:3, Isotope(139354187, Element(139354187, S:N, AN:7))))"
15:59 egonw hehehe
15:59 egonw yeah, but that works for me...
15:59 egonw N.sp2
15:59 slyrus right
15:59 egonw that is correct
15:59 egonw ok, got to run now
16:00 egonw can you send me an email, as reminder...
16:00 slyrus will do
16:00 egonw this structure makes a nice test case to see why the incorrect valence for C is causing trouble
16:00 egonw thanx
16:00 egonw cu later
17:52 egonw joined #cdk
19:27 jbrefort joined #cdk

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