#create table in sql

con = db.connection()

cur = db.cursor()

d = 'drop table if exists med_pages;'

q = 'create table med_pages(page_id bigint, title varchar, summary varchar, categories varchar, content text);'

for q in [d, q]:

cur.execute(q)

#load up desired terms from csv file

med_terms = []

with open('snomed_cleaned_term.txt','rb') as f:

text = f.readlines()

for line in text:

med_terms.extend(line.split(','))

con = db.connection()

cur = db.cursor()

missed = 0

i = 0

for term in med_terms:

#look in wikipedia for page associated with term

try:

page = wikipedia.page(term)

except wikipedia.exceptions.DisambiguationError as e:

#handle disambiguation error

e.options = [t.encode('utf-8') for t in e.options]

#prioritize choice if it has "medic" in title, grabs things like "(medicine)"

possible = [x for x in e.options if re.search('medic', x.lower())]

if possible:

try:

page = wikipedia.page(possible[0])

except:

missed += 1

continue

#otherwise take the first choice of term

else:

try:

page = wikipedia.page(e.options[0])

except:

missed += 1

continue

#if can't find any pages, skip term

except wikipedia.exceptions.PageError:

missed += 1

continue

#join all the categories by "," to make it a string for input into sql

try:

categories = ",".join(page.categories)

except:

categories = ""

#insert page into sql table

cur.execute("insert into med_pages VALUES (%s, %s, %s, %s, %s)",(int(page.pageid),page.title,page.summary,categories,page.content))

i += 1

con.commit()

print '# unidentifiable pages:', missed