We are no longer offering accounts on this server. Consider https://gitlab.freedesktop.org/ as a place to host projects.

Commit d079a33c authored by jarkko's avatar jarkko

Used generators to print data as we receive it.

parent 4e1d4ee9
......@@ -11,13 +11,11 @@ sys.stdout = codecs.lookup('utf-8')[-1](sys.stdout)
def parse_page(page):
soup = BeautifulSoup(urllib2.urlopen(page))
page_data = []
for row in soup.find('table', 'tracklist big').findAll('tr'):
artist, track, timestamp = parse_track(row)
# Tracks submitted before 2005 have no timestamp
if artist and track:
page_data.append((artist, track, timestamp))
return page_data
yield (artist, track, timestamp)
def parse_track(row):
......@@ -38,14 +36,12 @@ def fetch_tracks(user, request_delay=0.5):
num_pages = int(soup.find('a', 'lastpage').contents[0])
num_pages = 1
all_data = []
for cur_page in range(1, num_pages + 1):
data = parse_page(url + '?page=' + str(cur_page))
all_data += data
tracks = parse_page(url + '?page=' + str(cur_page))
for artist, track, timestamp in tracks:
yield (artist, track, timestamp)
if cur_page < num_pages:
return all_data
def main(*args):
if len(args) == 2:
Markdown is supported
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment