GCs GEO extraction

- #!/bin/python2.7
-out_labels = {
- '</Air>': 'O',
- '</Gtype>': 'O',
- '</Gversion>': 'O',
- '</Med>': 'O',
- '</Orgn>': 'O',
- '</Phase>': 'O',
- '</Sample>': 'O',
- '</Serie>': 'O',
- '</Strain>': 'O',
- '</Substrain>': 'O',
- '</Supp>': 'O',
- '</Technique>': 'O',
- '</Temp>': 'O',
- '</Name>': 'O',
- '</OD>': 'O',
- '</Anti>': 'O',
- '</Agit>': 'O',
- '</Vess>': 'O'}
-in_labels = {
- '<Air>': 'Air',
- '<Gtype>': 'Gtype',
- '<Gversion>': 'Gversion',
- '<Med>': 'Med',
- '<Orgn>': 'Orgn',
- '<Phase>': 'Phase',
- '<Sample>': 'Sample',
- '<Serie>': 'Serie',
- '<Strain>': 'Strain',
- '<Substrain>': 'Substrain',
- '<Supp>': 'Supp',
- '<Technique>': 'Technique',
- '<Temp>': 'Temp',
- '<Name>': 'Name',
- '<OD>': 'OD',
- '<Anti>': 'Anti',
- '<Agit>': 'Agit',
- '<Vess>': 'Vess'}
-
-import re
-#columna Contenido de "/home/egaytan/Dropbox/PGC/data-sets/file_output/exit_file.txt"
-inpath = '/home/egaytan/Dropbox/PGC/data-sets_1/content_colum_data_set.tsv.conll'
-outpath = '/home/egaytan/Dropbox/PGC/data-sets_1/sentences_labeled_v1.tsv'
-flag = 'O'
-with open(outpath, 'w') as out:
-	with open(inpath, 'r') as input_file:
-		for line in input_file:
-			if len(line.split('\t')) > 1:
-				w = line.split('\t')[1]
-				if w in in_labels or w in out_labels:
-					if w in in_labels.keys(): flag = in_labels[w]
-					if w in out_labels: flag = out_labels[w]
-
-				else:
-					if w == "PGCGROWTHCONDITIONS":	out.write('\n')
-					else: 
-						out.write('|'.join(line.split('\t')[1:4])+'|'+flag+' ')
-						#print('\t'.join(line.split('\t')[1:4])+'\t'+flag)
\ No newline at end of file