def wrapper(istring, annotator, index):
cur_dir = os.getcwd()
batch_path = cur_dir + r'\pretag\pretag.bat' # this is the batch file
print batch_path
os.chdir(".\data_final")
q_id = istring
q_raw_content = annotator + q_id + '-raw.txt'
q_tokenize = annotator + q_id + '-tokenized.txt'
q_transformin = annotator + q_id + '-input2tagger.txt'
q_pretag = annotator + q_id + '-pretagged.txt'
q_final = annotator + q_id + '-final.txt'
readSQL.read_knowledge_unit(q_id, q_raw_content)
mytokenizer.tokenize(q_raw_content, q_tokenize)
transforminput.transforminput(q_tokenize, q_transformin)
os.chdir("..\pretag")
child = subprocess.Popen([batch_path, "..\\data_final\\" + q_transformin, "..\\data_final\\" + q_pretag])
stdout, stderr = child.communicate()
child.wait() # wait() is important.
print("parent process")
os.chdir("..\data_final")
transformoutput.transformoutput(q_id, q_pretag, q_final)
shutil.copy2(q_final, annotator+str(index)+'.automatic_tags')
os.chdir("..")
print("program finish")
cur.execute("SELECT Id FROM posts where ParentId=%s" %(question_id))
answers = cur.fetchall()
for row in answers:
answer_id = row[0]
cur.execute("SELECT Body FROM posts where Id=%s" %(answer_id))
ans_body = cur.fetchall()
body += ans_body
cur.execute("SELECT Text FROM comments where PostId=%s" %(answer_id))
ans_comments = cur.fetchall()
comments += ans_comments
uni = ''.join( my_encoder( title[0][0] ) )
f.write(uni + '\n')
for row in body:
#uni = unicode(strip_tags(row[0]), 'utf-8', errors='replace_against_space')
uni = ''.join( my_encoder( strip_tags(''.join(my_encoder(row[0]) ) ) ) )
f.write(uni + '\n')
for row in comments:
uni = ''.join( my_encoder( strip_backtick(row[0]) ) )
f.write(uni + '\n')
f.close()
if __name__=='__main__':
# html = '''asdfasdf'''
# print strip_tags(''.join(my_encoder(html)))
dump('raw.txt') # dump to input_brown.txt
mytokenizer.tokenize('raw.txt', 'tokenized.txt')