Commit Graph
123 Commits
Author SHA1 Message Date
James Taylor 0ac85dac80 Commenting out another biomart test, they just hang forever. 2007-03-08 18:35:05 +00:00
James Taylor a70768b255 Buildbot functional test script updates. 2007-03-08 18:27:00 +00:00
James Taylor 95698117eb Unit test fix. 2007-03-08 18:25:01 +00:00
James Taylor cf5405b244 Bugfix. 2007-03-08 18:23:50 +00:00
James Taylor ea9a819235 Working on getting tests going again. Modified the history.xml template to
only return active datasets so deletes should test correctly again.
2007-03-08 18:14:50 +00:00
James Taylor 9370ea6a48 Don't echo database queries. 2007-03-08 18:11:51 +00:00
James Taylor 165cfdaf63 Remove psycopg dependency. 2007-03-08 17:58:16 +00:00
James Taylor 113bff99bc Use an association object for the relationship between parent and child
datasets.

THIS REQUIRES DATABASE CHANGES!

Also, test parse errors should now cause errors.
2007-03-08 17:52:29 +00:00
James Taylor e20cf00c91 "User" relation interface: normalizing form styles and such to be consistent
with the rest of the application.
2007-03-08 17:31:28 +00:00
James Taylor 4ac180091e Cheetah changes: upgrade to Cheetah 2.0 (release candidate 7, looking very
stable at this point). This gives us big improvements to template compilation
and caching, and thus a nice little speed boost.
2007-03-08 17:29:58 +00:00
Daniel Blankenberg d389d70b5b Add sequence to tablebrowser direct connection. Custom track can be added after an actual datatype is created. 2007-03-08 16:42:48 +00:00
Daniel Blankenberg 1ebbc0c583 Fix for hyperlinks for UCSC table browser direct connection. 2007-03-08 16:35:58 +00:00
David C. King 078389e8f0 aggregate_binned_scores now invokes the up-to-date tool 2007-03-07 22:49:54 +00:00
James Taylor 553b2d4c7b Made the conversion to HTML much prettier. 2007-03-05 22:39:06 +00:00
James Taylor 72260d2887 New profile middleware.
This resolved a lot of the problems with the profile middleware from Paste. In
particular:

  - Uses cProfile from Python 2.5, for Python 2.4 this is part of Armin Rigo's
    lsprof package (http://codespeak.net/svn/user/arigo/hack/misc/lsprof/)

  - Does not use redirect stdout, instead the Stats object is extended (in an
    unavoidably hacky way) to generate HTML output that gets embedded in the
    page.

  - Profile output can be viewed inline or in a popup window.

  - Doesn't require a lock, should be thread safe (but let me know if you see
    anything odd).

To use, just add 'use_profile = True' in the config file. Not for production
use, will slow things down, but no nearly as much as hotshot was.
2007-03-05 20:34:43 +00:00
Daniel Blankenberg 6e951364f8 Fix some broken logging. 2007-03-01 16:23:06 +00:00
Daniel Blankenberg 84e2dc5ed5 Fix for UCSC direct connection for unknown datatypes. 2007-03-01 16:22:29 +00:00
Daniel Blankenberg e013dc254e Make BED default datatype for UCSC direct connection. 2007-03-01 15:53:03 +00:00
Ian Schenck bdb37ba8ce Another change to ops to hopefully kill the never-ending strand problem. This makes anything that isn't a +/- into a +. 2007-02-28 21:59:38 +00:00
Ian Schenck f4f9e1d7b2 Preliminary data validation comit, much more to come. There will be a new datatype for errors which will hold pickled exceptions and repr them for display. In addition, errors should probably be a reserved or somehow special child dataset, so that there can be only one. Finally, implementing a tool or interface to fix errors needs to be done (and is why the exceptions must be pickled). Also, only the validation for intervals is done, as the parsers in bx for mafs, fasta, axt's, etc do not have any kind of validation built. 2007-02-28 21:46:54 +00:00
Daniel Blankenberg 412b9a9677 Add UCSC Table Browser direct connection tool. 2007-02-28 20:46:06 +00:00
Greg Von Kuster 0dfd869918 Enhanced column_maker.py to gracefully handle invalid data in input file and to display more meaningful info to user. 2007-02-28 16:10:31 +00:00
Greg Von Kuster 7ab5369033 Corrected the way these tools read input datasets to keep them from entering infinite loops if the dataset is empty. 2007-02-27 16:32:05 +00:00
Anton Nekrutenko 8c09569e13 Unsequenced genome for March form Dulles airport 2007-02-27 02:34:21 +00:00
Greg Von Kuster 0191564e9b Added another example since I noticed several jobs where the user was confused about the "==" requirement. 2007-02-26 20:43:31 +00:00
Greg Von Kuster 97e5cb8f9d Enhanced event logging for session activity reports. 2007-02-26 19:03:37 +00:00
Daniel Blankenberg e830b83b9f Removing import psyco_full from tools that throw warning (already imported).
Also, I noticed that the 2 aggregate score tools had the same tool id, and are live on the test server, so i fixed the tool ids.
2007-02-23 14:33:03 +00:00
Greg Von Kuster 5a8135f56c Minor correction to the message for the location of the first invalid line in a dataset, if there is in fact one. 2007-02-21 21:44:22 +00:00
James Taylor 7285237f5b Hack for database monitoring job queue to wait until after forking to start
monitor thread.
2007-02-21 20:16:35 +00:00
Greg Von Kuster 67e1bb8846 Enhanced the Filter tool to handle invalid data and to display more useful, informative messages to the user. 2007-02-21 18:41:58 +00:00
Daniel Blankenberg 573e8cd8a2 Parameter sanitizer can now be turned off by individual tools. Refresh behavior after tool execution can now be defined in
the tool xml file.

To Do: Allow individual parameters to have sanitation control.
2007-02-21 18:01:07 +00:00
Guruprasad Anada 9903924f97 Name change: Gene BED to Exon BED changed to Gene BED to Exon/Intron BED 2007-02-19 21:14:03 +00:00
Guruprasad Anada 84912a92fa Filename test_kaks.py changed to kaks_yn00.py 2007-02-19 19:34:26 +00:00
Guruprasad Anada be8f660cd3 yn00 binary added 2007-02-19 19:21:38 +00:00
Guruprasad Anada ac747b8d6f Modified tool: Gene BED to Exon BED 2007-02-19 16:59:51 +00:00
Guruprasad Anada 0ac0ec4086 "new tool: gene bed to intron bed- test output filr" 2007-02-19 16:25:54 +00:00
Guruprasad Anada 112f94c176 new tool: gene bed to intron bed 2007-02-19 16:22:28 +00:00
Guruprasad Anada fb3b9255f5 yn00 binary added: kaks tool 2007-02-16 21:37:34 +00:00
Guruprasad Anada b37f332a35 new toold kaks added 2007-02-16 21:28:21 +00:00
Nate Coraor 092c23f41b Made run.sh Bourne-friendly and added path fix for cluster nodes. 2007-02-16 21:21:24 +00:00
Daniel Blankenberg bfe9c45db9 Modify maf_to_bed to use database/tmp for its temp directory for additional output files. 2007-02-15 20:09:33 +00:00
Anton Nekrutenko e554fb4d79 added get_flanks plus two additional tools for a David Gilmore: srg2wig and aggTrack. These two are for demo only and should not go to main 2007-02-15 15:11:08 +00:00
James Taylor f6274675db Close connections after setting up the application (just is case the WSGI
server wants to fork, otherwise processes then try to share connections,
yikes!!!)
2007-02-13 23:49:49 +00:00
James Taylor 48cdd842f7 Support for the job "queue" reading new jobs from the database rather than
from a Queue.

Enable this when running the server in a forking environment. The job queue
threads will live in the parent process, while jobs will be added to the
database by the child processes. Lovely!
2007-02-13 21:00:18 +00:00
Nate Coraor 5c494af61a try the post-job hook so it doesn't kill the queue if it fails. 2007-02-13 17:54:43 +00:00
James Taylor 5a4182c243 1) A 'prefork threaded' mixin hacked together from existing flup components
2) An AJP server using that mixin (we could also do fastcgi or scgi, but AJP
   is the best option).
2007-02-13 17:27:25 +00:00
Guruprasad Anada 02a60145bf 2007-02-13 17:01:13 +00:00
Guruprasad Anada 0ef3bdbba0 modified test_kaks tool 2007-02-13 16:21:07 +00:00
James Taylor ed76bdaff2 The correct file for pre-fork server. 2007-02-13 05:17:32 +00:00
James Taylor bd0ff71748 rollback... 2007-02-13 05:17:18 +00:00