fixPipelineSetup-py

#!/usr/bin/python

import os

def get_immediate_subdirectories(dir):
    return [name for name in os.listdir(dir)
            if os.path.isdir(os.path.join(dir, name))]

# Get runfolder name from file and return as list.
def get_runfolders_already_run(file):
        runfolders = list()
        file = open(file)
        for line in file:
                runfolders.append(line.rstrip())
        return runfolders

def print_sample_folder_entries(runFolder):
        for dir in get_immediate_subdirectories(runFolder):
                if ("Sample_" in dir):
                        sampleName = dir.replace("Sample_","")
                        samplePath = runFolder + "/" + dir
                        print ("\t\t<SampleFolder Name=\"%s\" Path=\"%s/\" Reference=\"reference/Homo_sapiens.GRCh37.57.dna.concat.fa\"></SampleFolder>" % (sampleName, samplePath))


#------------------------------------------------------------------
# Make a list of all runfolders which have already been processed.
#------------------------------------------------------------------

# Read a file of with the runfolder names and add them to a hash
        #-> define function to get this hash

runFoldersAlreadyRunFile = "runFoldersAlreadyRun.txt"
runFoldersAlreadyRun = get_runfolders_already_run(runFoldersAlreadyRunFile)
rootDir = "../../"

#------------------------------------------------------------------
#
#------------------------------------------------------------------

print """<Project Name="Some project" SequencingCenter="SnqSeq - Uppsala"    Platform="Illumina" UppmaxProjectId="a2009002">"""

for runFolder in get_immediate_subdirectories(rootDir):
        if "_" in runFolder.lower() and runFolder not in runFoldersAlreadyRun:
                print("\t<RunFolder Report=\"%s%s/report.xml\">" % (rootDir, runFolder))
                print_sample_folder_entries(rootDir + runFolder)
                print("\t</RunFolder>")
                with open(runFoldersAlreadyRunFile,"a") as f:
                        f.write("\n" + runFolder)
print "</Project>"