def read():
db = DataBase('not_bsd.dat')
a = db['a']
b = db['b']
db.close()
print('a', a)
print('b', b)
def load_workers(self, filename):
"""
Return a list of workers read from disk as [(id, started, assetid),...].
"""
shelf = DbfilenameShelf(filename)
try:
workers = shelf['workers']
except:
workers = []
shelf.close()
return workers
def dump_workers(self, filename, workers):
"""
Write a sequence of workers written to disk, e.g.
[(id, started, assetid),...], and then return the sequence.
"""
seq = []
for w in workers:
seq.append((w['worker'], w['started'], w['args'][1]))
shelf = DbfilenameShelf(filename)
shelf['workers'] = seq
shelf.close()
return seq
def update_shelf():
url_mention = DbfilenameShelf(args.in_shelf, protocol=-1)
TM = url_mention['__TOKEN_MAPPER__']
TM.finalize(catpeople_baseline_nb_config.MAX_TOK)
E = url_mention['__URL_LIST__']
n_doc = 10000
with rasengan.tictoc('Extracting Contexts'):
df_obj = TextualClueObject(E[:n_doc], url_mention, TM)
df = defaultdict(int)
for features in df_obj.features.itervalues():
for f in features:
df[f] += 1
for f in df.keys():
df[f] = df[f] / float(n_doc)
url_mention['__DF__'] = dict(df)
url_mention.close()
return
del seg[e_sent_tok_idx]
else:
# import pdb; pdb.set_trace()
print >> sys.stderr, "Skipped mention", mention[1]
continue
if seg[e_sent][1] - seg[e_sent][0] > MAX_CHAR_IN_SENT:
# The sentence that contains the entity is too long.
print >> sys.stderr, '__noisy__', text[
seg[e_sent][0]:seg[e_sent][1]]
continue
e_start = mention_start - seg[e_sent][0]
# Remove too long sentences.
sentences = [[tmp_tok.lower() for tmp_tok in PAT_TOKENIZER(text[a:b])]
for (a, b) in seg if b - a <= MAX_CHAR_IN_SENT]
mapped_sentences = [TOKEN_MAPPER(e) for e in sentences]
# Adjust pointer to sentence that contain the entity since we
# might have removed some extremely long sentences.
# idx_to_sentence_that_contains_entity
itste = (e_sent -
sum(1 for (a, b) in seg[:e_sent] if b - a > MAX_CHAR_IN_SENT))
esidx, eeidx = tokens_in_tokenization_corresponding_to_a_span(
text[seg[itste][0]:seg[itste][1]], e_start, e_start + len_mention,
sentences[itste])
out_mention = [mapped_sentences, itste, esidx, eeidx]
out_mentions.append(out_mention)
pass
out_shelf[url] = deduplicate_unhashables(out_mentions)
out_shelf['__URL_LIST__'] = urls
out_shelf['__TOKEN_MAPPER__'] = TOKEN_MAPPER
out_shelf.close()
connection.commit()
session_store['authenticated'] = False
form = ''
message = '<p>Your account has been successfully deleted.</p>'
cursor.close()
connection.close()
else:
message = '<p><strong>Error! Passwords must match.</strong></p>'
else:
message = '<p><strong>Error! Please check the box to confirm deletion.</strong></p>'
result = """<section>
<h2>Delete Account</h2>"""
result += form
result += message
result += '</section>'
session_store.close()
except (db.Error, IOError):
result = '<p>Sorry! We are experiencing problems at the moment. Please call back later.</p>'
print("""
<!DOCTYPE html>
<html lang="en">
<head>
<meta charset="utf-8" />
<link rel="stylesheet" href="styles.css" />
<meta name="viewport" content="initial-scale=1.0, width=device-width" />
<title>Delete Account</title>
</head>
<body>
<header>
<h1>a-MAZE-ing</h1>
out_val.default_factory = None # FINALIZE out_val
return out_val
with rasengan.tictoc('Reading wikilinks'):
# With joblib this takes only 8 minutes !!
from joblib import Parallel, delayed
out_val_list = Parallel(n_jobs=10)(
delayed(get_mention_from_wikilink_thrift_file)(fn)
for fn in range(1, args.last_f2r))
# out_val_list = [get_mention_from_wikilink_thrift_file(fn)
# for fn in range(1, args.last_f2r)]
with rasengan.tictoc('Shelving'):
import shelve
from shelve import DbfilenameShelf
total_data = defaultdict(list)
for out_val in out_val_list:
for url in out_val:
total_data[url].extend(out_val[url])
total_data.default_factory = None # FINALIZE out_val
# Save the results of the processing.
shelf = DbfilenameShelf(args.out_fn, protocol=-1)
shelf['__URL_LIST__'] = total_data.keys()
for url in shelf['__URL_LIST__']:
shelf[url] = total_data[url]
shelf.close()
# Validation
for e in POOL:
assert e in total_data