diff validator.py @ 0:01ed8e112f2a draft

planemo upload commit 6bf6d4ec8ff2ec6bd45b5f483ff2f83b6229d57d
author yating-l
date Wed, 12 Apr 2017 17:44:58 -0400
parents
children
line wrap: on
line diff
--- /dev/null	Thu Jan 01 00:00:00 1970 +0000
+++ b/validator.py	Wed Apr 12 17:44:58 2017 -0400
@@ -0,0 +1,39 @@
+import sys
+
+"""
+Call checkAndFixBed, check the integrity of bed file. If the strand is not "+" or "-" truncate that line and report to users
+create a column and move the score column to that column.
+"""
+def checkAndFixBed(bedfile, revised_file):
+# Store the lines that have been removed
+    removedLines = []
+# Remove the lines with invalid strand, create a score column to store the original scores and change scores in the original score column all to 1000
+    with open(revised_file, 'w') as tmp:
+        with open(bedfile, 'r') as f:
+            lines = f.readlines()
+            i = 1
+            for line in lines:  
+                fields = line.split()
+                strand = fields[5]
+                score = fields[4]
+                fields[4] = '1000' 
+                fields.append(score)
+                if (strand == '+' or strand == '-'):
+                    tmp.write('\t'.join(map(str, fields)))
+                    tmp.write("\n")
+                else:
+                    removedLines.append("line" + str(i) + ": " + line)
+                i = i+1
+
+    return removedLines
+
+def main():
+    inputfile = str(sys.argv[1])
+    outputfile = str(sys.argv[2])
+    removed = checkAndFixBed(inputfile, outputfile)
+    if (removed != []):
+        print "\nRemoved invalid lines: \n"
+        print "\n".join(removed)
+
+if __name__ == "__main__":
+    main()