Mercurial > repos > marcel > caddsuite_linux_x86_64
diff CADDSuite/galaxyconfigs/molFiles.py @ 3:ff6df146c0d5
Uploaded
author | marcel |
---|---|
date | Sun, 26 Jun 2011 13:53:57 -0400 |
parents | bac3c274238f |
children |
line wrap: on
line diff
--- /dev/null Thu Jan 01 00:00:00 1970 +0000 +++ b/CADDSuite/galaxyconfigs/molFiles.py Sun Jun 26 13:53:57 2011 -0400 @@ -0,0 +1,128 @@ +# -*- coding: utf-8 -*- + +import data +import logging +from galaxy.datatypes.sniff import * +import commands + +log = logging.getLogger(__name__) + +class GenericMolFile( data.Text ): + file_ext = "mol2/sdf/drf" + def check_filetype( self,filename ): + self.no_mols = commands.getstatusoutput("grep -c \\$\\$\\$\\$ "+filename) + if (self.no_mols[0] == 0) & (self.no_mols[1] > 0): + self.file_ext="sdf" + return True + self.no_mols = commands.getstatusoutput("grep -c @\<TRIPOS\>MOLECULE "+filename) + if (self.no_mols[0] == 0) & (self.no_mols[1] > 0): + self.file_ext="mol2" + return True + self.no_mols = commands.getstatusoutput("grep -c \"ligand id\" "+filename) + if (self.no_mols[0] == 0) & (self.no_mols[1] > 0): + self.file_ext="drf" + return True + self.no_mols = commands.getstatusoutput("grep -c HEADER "+filename) + if (self.no_mols[0] == 0) & (self.no_mols[1] > 0): + self.file_ext="pdb" + return True + return False + + def set_peek( self, dataset, is_multi_byte=False ): + if not dataset.dataset.purged: + if(self.check_filetype(dataset.file_name)) : + if (self.no_mols[1] == '1'): + dataset.blurb = "1 molecule" + else: + dataset.blurb = "%s molecules"%self.no_mols[1] + dataset.peek = data.get_file_peek( dataset.file_name, is_multi_byte=is_multi_byte ) + else: + dataset.peek = 'file does not exist' + dataset.blurb = 'file purged from disk' + + def get_mime(self): + return 'text/plain' + + +class GenericMultiMolFile( GenericMolFile ): + def set_peek( self, dataset, is_multi_byte=False ): + if not dataset.dataset.purged: + self.sniff(dataset.file_name) + if (self.no_mols[1] == '1'): + dataset.blurb = "1 molecule" + else: + dataset.blurb = "%s molecules"%self.no_mols[1] + dataset.peek = data.get_file_peek( dataset.file_name, is_multi_byte=is_multi_byte ) + else: + dataset.peek = 'file does not exist' + dataset.blurb = 'file purged from disk' + +class SDF( GenericMultiMolFile ): + file_ext = "sdf" + def sniff( self, filename ): + self.no_mols = commands.getstatusoutput("grep -c \\$\\$\\$\\$ "+filename) + if (self.no_mols[0] == 0) & (self.no_mols[1] > 0): + return True + else: + return False + +class MOL2( GenericMultiMolFile ): + file_ext = "mol2" + def sniff( self, filename ): + self.no_mols = commands.getstatusoutput("grep -c @\<TRIPOS\>MOLECULE "+filename) + if (self.no_mols[0] == 0) & (self.no_mols[1] > 0): + return True + else: + return False + +class DRF( GenericMultiMolFile ): + file_ext = "drf" + def sniff( self, filename ): + self.no_mols = commands.getstatusoutput("grep -c \"ligand id\" "+filename) + if (self.no_mols[0] == 0) & (self.no_mols[1] > 0): + return True + else: + return False + +class PDB( GenericMolFile ): + file_ext = "pdb" + def sniff( self, filename ): + self.no_mols = commands.getstatusoutput("grep -c HEADER "+filename) + if (self.no_mols[0] == 0) & (self.no_mols[1] > 0): + return True + else: + return False + def set_peek( self, dataset, is_multi_byte=False ): + #def set_peek( self, dataset, line_count=None, is_multi_byte=False ): + if not dataset.dataset.purged: + res = commands.getstatusoutput("lib/galaxy/datatypes/countResidues.sh "+dataset.file_name) + dataset.peek = res[1] + self.sniff(dataset.file_name) + if (self.no_mols[1] == '1'): + dataset.blurb = "1 protein structure" + else: + dataset.blurb = "%s protein structures"%self.no_mols[1] + else: + dataset.peek = 'file does not exist' + dataset.blurb = 'file purged from disk' + +class grd ( data.Text ) : + file_ext = "grd" + def set_peek( self, dataset, is_multi_byte=False ): + if not dataset.dataset.purged: + #dataset.peek = "" + dataset.blurb = "score-grids for docking" + else: + dataset.peek = 'file does not exist' + dataset.blurb = 'file purged from disk' + +class grdtgz ( data.Text ) : + file_ext = "grd.tgz" + def set_peek( self, dataset, is_multi_byte=False ): + if not dataset.dataset.purged: + #dataset.peek = "" + dataset.blurb = "compressed score-grids for docking" + else: + dataset.peek = 'file does not exist' + dataset.blurb = 'file purged from disk' +