def add_tags(_, update, args, quoted_sticker_id):
message = update.message
arg_str = ' '.join(args)
tags = extract_tags(arg_str)
if not tags:
message.reply_text('Usage: /add_tags *tags*')
return
if not sticker_storage.has_sticker_with_file_id(quoted_sticker_id):
sticker_storage.add(file_id=quoted_sticker_id,
owner_id=message.from_user.id)
old_tags = sticker_storage.get_by_file_id(file_id=quoted_sticker_id,
tagged=True).tags
sticker_storage.add_tags_by_file_id(file_id=quoted_sticker_id,
tags=tags,
owner_id=message.from_user.id)
current_tags = sticker_storage.get_by_file_id(file_id=quoted_sticker_id,
tagged=True).tags
added_tags = set(current_tags) - set(old_tags)
if added_tags:
result_message = ('Tags have been added!\n\n' + 'New tags: ' +
', '.join(sorted(added_tags)))
else:
result_message = 'No new tags have been added.'
message.reply_text(result_message + '\n\n' + 'Current tags: ' +
', '.join(sorted(current_tags)))
def blog():
""" blog page """
content = u.get_content()["blog"]
# Get all possible filtering tags
tags = u.extract_tags(content.values())
return render_template(
"blog.html", html_tab_title="Villoro - Blog", blog=content, tags=tags, **u.get_page("blog")
)
def portfolio():
""" portfolio page """
content = u.get_content()["portfolio"]
# Get all possible filtering tags
tags = u.extract_tags(content.values())
return render_template("portfolio.html",
portfolio=content,
tags=tags,
**u.get_page("portfolio"))
def tags_handler(_, update, user_data):
message = update.message
text = message.text.strip()
tags = extract_tags(text)
sticker_id = sticker_storage.add(file_id=user_data['sticker_file_id'],
owner_id=message.from_user.id)
sticker_storage.add_tags(sticker_id=sticker_id,
tags=tags,
owner_id=message.from_user.id)
message.reply_text('Sticker has been added.')
return ConversationHandler.END
return attempt
raise Exception("Not found: " + fname)
if output == "imagesamples":
if dataset == 'ukbench':
groupsize = 4
candidate_id = random.sample(range(len(image_names)), 1)[0]
selected = extract_fname_id_test(image_names[candidate_id])
base_id = int(selected / groupsize) * groupsize
targets = ["ukbench%05d.jpg" % i for i in range(base_id, base_id + groupsize)]
captions = None
elif dataset == 'mirflickr':
training_images = utils.read_image_tags("../datasets/mirflickr", "-training")
target_id = random.sample(range(len(image_names) / 2, len(image_names)), 3)[0]
target = image_names[target_id]
matches = utils.retrieve_matches(utils.extract_tags(target), training_images)
targets = []
targets.append(target)
for i in range(3):
targets.append(matches[i])
captions = {}
for target in targets:
captions[target] = ", ".join(utils.extract_tags(target))
else:
raise Exception('Unsupported')
print(tex_image_table(targets, captions=captions))
if output in ["results1", "results2"]:
if dataset == 'ukbench':
captions = None
if output == "results1":
def main():
vbmap = load_vbmap(sys.argv[1])
masters = extract_masters(vbmap['map'])
replicas = extract_replicas(vbmap['map'])
nodes = extract_nodes(vbmap['map'])
nodes_count = len(nodes)
nodes_dict = dict((n, i) for i, n in enumerate(nodes))
tags = extract_tags(vbmap, nodes)
tags_list = sorted(set(tags.values()))
tags_count = len(tags_list)
pylab.figure()
pylab.subplot(211)
pylab.xticks([i + 0.5 for i in xrange(nodes_count)], nodes)
plots = [hist(masters, nodes_dict)] + \
[hist(r, nodes_dict) for r in replicas]
labels = ['master'] + ['replica %d' % i for i in xrange(len(replicas))]
pylab.hist(plots, bins=xrange(nodes_count + 1), label=labels)
pylab.title("Number of vbuckets per node")
pylab.xlabel("Nodes")
pylab.ylabel("Number of vbuckets")
pylab.legend()
pylab.subplot(212)
pylab.xticks([i + 0.5 for i in xrange(nodes_count)], nodes)
all_replicas = list(chain(*replicas))
pylab.hist(hist(all_replicas, nodes_dict), bins=xrange(nodes_count + 1),
label='all replicas', rwidth=0.5)
pylab.title("Number of replica vbuckets per node")
pylab.xlabel("Nodes")
pylab.ylabel("Number of vbuckets")
pylab.legend()
pylab.figure()
pylab.subplot(211)
plots = [[tags[n] for n in masters]] + \
[[tags[n] for n in r] for r in replicas]
pylab.hist(plots, bins=xrange(tags_count + 1), label=labels)
pylab.xticks([i + 0.5 for i in xrange(tags_count)], tags_list)
pylab.title("Number of vbuckets per tag")
pylab.xlabel("Tags")
pylab.ylabel("Number of vbuckets")
pylab.legend()
pylab.subplot(212)
pylab.xticks([i + 0.5 for i in xrange(nodes_count)], nodes)
pylab.title("Number of nodes each node replicates to per tag")
pylab.xlabel("Nodes")
pylab.ylabel("Number of replica nodes")
tags_repcounts = tag_replication_counts(vbmap['map'], nodes, tags_list, tags)
plots = []
for tag_counts in tags_repcounts:
plot = []
for node, count in enumerate(tag_counts):
plot.extend([node] * count)
plots.append(plot)
pylab.hist(plots, bins=xrange(nodes_count + 1), label=map(str, tags))
pylab.legend()
simulate(vbmap)
pylab.show()
def test_that_extract_tags_leaves_only_alphanumeric_and_underline_characters():
assert extract_tags('this.is.a test') == ['thisisa', 'test']
def test_that_extract_tags_ignores_trailing_whitespace_and_comma_characters():
assert extract_tags('this is a test,') == ['this', 'is', 'a', 'test']
assert extract_tags('this is a test,,') == ['this', 'is', 'a', 'test']
assert extract_tags('this is a test ') == ['this', 'is', 'a', 'test']
assert extract_tags('this is a test ') == ['this', 'is', 'a', 'test']
assert extract_tags(', this is a test, ') == ['this', 'is', 'a', 'test']
def test_that_extract_tags_splits_on_whitespace_and_commma_characters():
assert extract_tags('this is a test') == ['this', 'is', 'a', 'test']
assert extract_tags('this, is, a, test') == ['this', 'is', 'a', 'test']
assert extract_tags('this, is a test') == ['this', 'is', 'a', 'test']
assert extract_tags('this,,,,,is,a,test') == ['this', 'is', 'a', 'test']
def generate_graphics_statement(image_path, width='.5in'):
return "\\includegraphics[keepaspectratio=false, height=%s, width=%s]{%s}" % (width, width, image_path)
targets = {}
current_tags = None
for line in lines1:
if line.startswith('src'):
continue
if 'datasets' in line:
if line.startswith('datasets'):
current = line.split()[0]
targets[current] = []
current_tags = extract_tags(current)
else:
image_path = line.split()[1]
if len(targets[current]) >= 4:
continue
tags = extract_tags(image_path)
targets[current].append((image_path, set(tags).intersection(set(current_tags))))
from pprint import pprint
# fix up self
for k, v in targets.iteritems():
others = []
for i in range(1, 4):
image_path, tags = targets[k][i]
others.extend(tags)
def extract_tags(text, topK=10):
tf_tags = utils.extract_tags(text, topK=topK)
rank_tags = utils.textrank(text, topK=topK)
# 取tfidf和textrank结果的并集
return set(tf_tags + rank_tags)
