def comment_info(comment):
comment_name = u'comment_{0}'.format(comment['id'])
if not Entity.by_name(comment_name):
print("Caching new comment {0}".format(comment_name))
entity = Entity(comment_name)
entity[u'body'] = comment['body']
DBSession.add(entity)
DBSession.commit()
return Entity.by_name(comment_name)
def comment_info(comment):
comment_name = u'comment_{0}'.format(comment['id'])
if not Entity.by_name(comment_name):
print("Caching new comment {0}".format(comment_name))
entity = Entity(comment_name)
entity[u'body'] = comment['body']
DBSession.add(entity)
DBSession.commit()
return Entity.by_name(comment_name)
def issue_info(issue):
issue_name = u'issue_{0}'.format(issue['id'])
if not Entity.by_name(issue_name):
print("Caching new issue {0}".format(issue_name))
entity = Entity(issue_name)
entity[u'title'] = issue['title']
entity[u'number'] = issue['number']
DBSession.add(entity)
DBSession.commit()
return Entity.by_name(issue_name)
def issue_info(issue):
issue_name = u'issue_{0}'.format(issue['id'])
if not Entity.by_name(issue_name):
print("Caching new issue {0}".format(issue_name))
entity = Entity(issue_name)
entity[u'title'] = issue['title']
entity[u'number'] = issue['number']
DBSession.add(entity)
DBSession.commit()
return Entity.by_name(issue_name)
def build_repo(self, repo_id, repo_info):
repo = Entity.by_name(repo_id)
owner = Entity.by_name(repo['owner'])
text = ["{0} is a cool project!".format(repo['name'])]
for user, count in repo_info.items():
if user == 'count':
continue
text.append("{0} made {1} contributions this week.".format(
Entity.by_name(user)['name'], count))
self.finish(owner, text)
def build_repo(self, repo_id, repo_info):
repo = Entity.by_name(repo_id)
owner = Entity.by_name(repo['owner'])
text = ["{0} is a cool project!".format(repo['name'])]
for user, count in repo_info.items():
if user == 'count':
continue
text.append("{0} made {1} contributions this week."
.format(Entity.by_name(user)['name'], count))
self.finish(owner, text)
def inject_knowledge():
#SunlightAPI returns nicely Unicoded data, but not all APIs will...
#Don't forget to use to_unicode()
knowledge = DBSession
for leg in all_legs:
print leg['full_name'], leg['state']
if Entity.by_name(leg['full_name']):
continue
character = Entity.by_name(leg['full_name'])
else:
character = Entity(leg['full_name'])
for key, value in leg.items():
character[key] = to_unicode(value)
knowledge.add(character)
knowledge.commit()
def build_user(self, user_id, user_info):
user = Entity.by_name(user_id)
display_key = {
'CommitCommentEvent': ('commented on', 'commits'),
'CreateEvent': ('created', 'tags, branches, or repositories'),
'DeleteEvent': ('deleted', 'tags, branches, or repositories'),
'FollowEvent': ('followed', 'users'),
'ForkEvent': ('forked', 'repositories'),
'GistEvent': ('made or modified', 'gists'),
'GollumEvent': ('made or modified', 'wiki pages'),
'IssueCommentEvent': ('commented on', 'issues'),
'IssuesEvent': ('made or modified', 'issues'),
'PullRequestEvent': ('made or modified', 'pull requests'),
'PullRequestReviewCommentEvent': ('commented on', 'pull requests'),
'PushEvent': ('pushed', 'commits'),
'WatchEvent': ('watched', 'repositories'),
}
text = ["{0} has been very busy this week!".format(user['name'])]
for event_type, count in user_info.items():
if event_type == 'count':
continue
elif event_type in ['CommitCommentEvent', 'FollowEvent',
'IssueCommentEvent', 'WatchEvent',
'PullRequestReviewCommentEvent',]:
count = int(count * 10)
display_text = display_key.get(event_type, ('made', event_type))
text.append("{0} {1[0]} {2} {1[1]} this week."
.format(user['name'], display_text, count, event_type))
self.finish(user, text)
def top_contributions():
week_activity = recent_events(7)
user_activity = defaultdict(lambda: defaultdict(int))
repo_activity = defaultdict(lambda: defaultdict(int))
for event in week_activity:
changes = 1
if event['type'] == 'PushEvent':
changes = event['payload']['size']
elif event['type'] in ['CommitCommentEvent', 'FollowEvent',
'IssueCommentEvent', 'WatchEvent',
'PullRequestReviewCommentEvent',]:
# Social (non-coding) events carry less weight
changes = .1
user_activity[event['actor']]['count'] += changes
user_activity[event['actor']][event['type']] += changes
if Entity.by_name(event['repo']):
if event['type'] in ['CommitCommentEvent', 'FollowEvent',
'IssueCommentEvent', 'WatchEvent',
'PullRequestReviewCommentEvent',]:
changes = 1
repo_activity[event['repo']]['count'] += changes
repo_activity[event['repo']][event['actor']] += changes
return user_activity, repo_activity
def build_user(self, user_id, user_info):
user = Entity.by_name(user_id)
display_key = {
'CommitCommentEvent': ('commented on', 'commits'),
'CreateEvent': ('created', 'tags, branches, or repositories'),
'DeleteEvent': ('deleted', 'tags, branches, or repositories'),
'FollowEvent': ('followed', 'users'),
'ForkEvent': ('forked', 'repositories'),
'GistEvent': ('made or modified', 'gists'),
'GollumEvent': ('made or modified', 'wiki pages'),
'IssueCommentEvent': ('commented on', 'issues'),
'IssuesEvent': ('made or modified', 'issues'),
'PullRequestEvent': ('made or modified', 'pull requests'),
'PullRequestReviewCommentEvent': ('commented on', 'pull requests'),
'PushEvent': ('pushed', 'commits'),
'WatchEvent': ('watched', 'repositories'),
}
text = ["{0} has been very busy this week!".format(user['name'])]
for event_type, count in user_info.items():
if event_type == 'count':
continue
elif event_type in [
'CommitCommentEvent',
'FollowEvent',
'IssueCommentEvent',
'WatchEvent',
'PullRequestReviewCommentEvent',
]:
count = int(count * 10)
display_text = display_key.get(event_type, ('made', event_type))
text.append("{0} {1[0]} {2} {1[1]} this week.".format(
user['name'], display_text, count, event_type))
self.finish(user, text)
def top_contributions():
week_activity = recent_events(7)
user_activity = defaultdict(lambda: defaultdict(int))
repo_activity = defaultdict(lambda: defaultdict(int))
for event in week_activity:
changes = 1
if event['type'] == 'PushEvent':
changes = event['payload']['size']
elif event['type'] in [
'CommitCommentEvent',
'FollowEvent',
'IssueCommentEvent',
'WatchEvent',
'PullRequestReviewCommentEvent',
]:
# Social (non-coding) events carry less weight
changes = .1
user_activity[event['actor']]['count'] += changes
user_activity[event['actor']][event['type']] += changes
if Entity.by_name(event['repo']):
if event['type'] in [
'CommitCommentEvent',
'FollowEvent',
'IssueCommentEvent',
'WatchEvent',
'PullRequestReviewCommentEvent',
]:
changes = 1
repo_activity[event['repo']]['count'] += changes
repo_activity[event['repo']][event['actor']] += changes
return user_activity, repo_activity
def repo_info(repo):
repo_name = repo.get('full_name', '{0}/{1}'.format(repo['owner']['login'],
repo['name']))
if not Entity.by_name(repo_name):
print("Caching new repository {0}".format(repo_name))
entity = Entity(repo_name)
entity['name'] = repo['full_name']
# Evidently you cannot set facts to None. (?)
if not repo['description']:
entity['description'] = u''
else:
entity['description'] = repo['description']
entity['url'] = repo['html_url']
entity['owner'] = user_info(repo['owner']).name
DBSession.add(entity)
DBSession.commit()
return Entity.by_name(repo_name)
def user_info(user):
user_name = u'user_{0}'.format(user['id'])
if not Entity.by_name(user_name):
print("Caching new user {0}".format(user_name))
entity = Entity(user_name)
entity['login'] = user['login']
entity['gravatar'] = user['gravatar_id']
entity['avatar'] = u'http://www.gravatar.com/avatar/{0}?s=200' \
.format(user['gravatar_id'])
# Not everyone has set a name for their account.
if user.get('name'):
entity[u'name'] = user['name']
else:
entity[u'name'] = user['login']
DBSession.add(entity)
DBSession.commit()
return Entity.by_name(user_name)
def repo_info(repo):
repo_name = repo.get(
'full_name', '{0}/{1}'.format(repo['owner']['login'], repo['name']))
if not Entity.by_name(repo_name):
print("Caching new repository {0}".format(repo_name))
entity = Entity(repo_name)
entity['name'] = repo['full_name']
# Evidently you cannot set facts to None. (?)
if not repo['description']:
entity['description'] = u''
else:
entity['description'] = repo['description']
entity['url'] = repo['html_url']
entity['owner'] = user_info(repo['owner']).name
DBSession.add(entity)
DBSession.commit()
return Entity.by_name(repo_name)
def user_info(user):
user_name = u'user_{0}'.format(user['id'])
if not Entity.by_name(user_name):
print("Caching new user {0}".format(user_name))
entity = Entity(user_name)
entity['login'] = user['login']
entity['gravatar'] = user['gravatar_id']
entity['avatar'] = u'http://www.gravatar.com/avatar/{0}?s=200' \
.format(user['gravatar_id'])
# Not everyone has set a name for their account.
if user.get('name'):
entity[u'name'] = user['name']
else:
entity[u'name'] = user['login']
DBSession.add(entity)
DBSession.commit()
return Entity.by_name(user_name)
def inject_knowledge():
#SunlightAPI returns nicely Unicoded data, but not all APIs will...
#Don't forget to use to_unicode()
knowledge = DBSession
for leg in ny_legs:
print leg['full_name']
#Check if entity is already in DB, and if so continue
#This is good for not duplicating entries, however
#This will skip updates to the entity :(
#This is why Sunlight included an "updated_at" field :)
#TODO: Change check to "updated_at" instead of full_name
if Entity.by_name(leg['full_name']):
continue
character = Entity.by_name(leg['full_name'])
else:
character = Entity(leg['full_name'])
for key, value in leg.items():
character[key] = to_unicode(value)
knowledge.add(character)
knowledge.commit()
def event_info(event):
event_name = u'event_{0}'.format(event['id'])
if not Entity.by_name(event_name):
print("Caching new event {0}".format(event_name))
entity = Entity(event_name)
entity['name'] = event_name
entity[u'actor'] = user_info(event['actor']).name
try:
entity[u'repo'] = repo_info(event['repo']['name']).name
except:
entity['repo'] = event['repo']['name']
entity[u'type'] = event['type']
entity[u'payload'] = event['payload']
entity[u'created_at'] = datetime.strptime(event['created_at'], '%Y-%m-%dT%H:%M:%SZ')
if 'Comment' in event['type']:
entity[u'comment'] = comment_info(event['payload']['comment']).name
if 'Issue' in event['type']:
entity['issue'] = issue_info(event['payload']['issue']).name
DBSession.add(entity)
DBSession.commit()
return Entity.by_name(event_name)
def event_info(event):
event_name = u'event_{0}'.format(event['id'])
if not Entity.by_name(event_name):
print("Caching new event {0}".format(event_name))
entity = Entity(event_name)
entity['name'] = event_name
entity[u'actor'] = user_info(event['actor']).name
try:
entity[u'repo'] = repo_info(event['repo']['name']).name
except:
entity['repo'] = event['repo']['name']
entity[u'type'] = event['type']
entity[u'payload'] = event['payload']
entity[u'created_at'] = datetime.strptime(event['created_at'],
'%Y-%m-%dT%H:%M:%SZ')
if 'Comment' in event['type']:
entity[u'comment'] = comment_info(event['payload']['comment']).name
if 'Issue' in event['type']:
entity['issue'] = issue_info(event['payload']['issue']).name
DBSession.add(entity)
DBSession.commit()
return Entity.by_name(event_name)
def zip_exe_handler(self, entity):
""" Handles self-extracting zip files """
self.log.debug("zip_exe_handler(%s)" % entity)
entity[u'format'] = u'zip'
dirname = os.path.dirname(entity[u'filename'])
p = subprocess.Popen('unzip -o "%s"' % entity[u'filename'],
shell=True, stdout=subprocess.PIPE,
stderr=subprocess.PIPE,
cwd=dirname)
out, err = p.communicate()
if err:
self.log.error("Error unzipping: " + err)
else:
# Delete compressed data after extracting
os.unlink(entity[u'filename'])
for line in out.split('\n'):
if line.strip().startswith('inflating'):
extracted = os.path.join(dirname, line.strip().split()[-1])
self.log.debug("extracted " + extracted)
magic = utils.get_magic(extracted)
# Create a new child Entity for each extracted file
extracted = to_unicode(extracted)
child = Entity.by_name(extracted)
if not child:
child = Entity(name=os.path.basename(extracted))
child[u'filename'] = extracted
DBSession.add(child)
child.parent = entity
child[u'magic'] = to_unicode(magic)
self.log.debug("Created %s" % child)
else:
child.parent = entity
DBSession.flush()
self.call_magic_handler(extracted, child)
def cache_events(client, org):
"""Pull new events from Github."""
try:
members = client.organization_members(org)
except:
print('Error getting members')
return
for user in members:
try:
events = client.user_activity(user['login'])
for event in events:
if not Entity.by_name(event['repo']):
client.repo_information(event['repo'])
except:
print("Something went wrong updating the events for {0}." \
.format(user['login']))
continue
print("You have {0} of {1} calls left this hour."
.format(*client.rate_limiting))
def populate(self, event):
self.event = event
user = Entity.by_name(event[u'actor'])
user_name = user[u'name']
repo = event[u'repo']
if not Entity.by_name(repo):
repo_link = event[u'repo']
repo_desc = ''
else:
repo = Entity.by_name(repo)
repo_link = '<a href="{0}">{1}</a>'.format(repo['url'], repo['name'])
repo_desc = repo['description']
self.box.pack_start(url_to_image(user[u'avatar'], user[u'gravatar'], self.scale),
False, False, 10)
event_colors = {
'commit': "#C9FFC1",
'branch': "#C2C9FF",
'issue': "#FFBAF9",
'comment': "#FFDDBD",
'social': "#FFFF80",
}
event_text = []
color = "#FFFFFF"
if event[u'type'] == "CommitCommentEvent":
color = event_colors['comment']
event_text.append("{0} commented on a commit in {1}."
.format(user_name, repo_link))
comment = Entity.by_name(event[u'comment'])
event_text.append(comment[u'body'])
elif event[u'type'] == "CreateEvent":
color = event_colors['branch']
new_type = event[u'payload']['ref_type']
if new_type == 'repository':
event_text.append("{0} created a new {1}, {2}."
.format(user_name, new_type, repo_link))
else:
event_text.append("{0} created {1} <tt>{3}</tt> in {2}."
.format(user_name, new_type, repo_link,
event[u'payload']['ref']))
event_text.append(repo_desc)
elif event[u'type'] == "DeleteEvent":
color = event_colors['branch']
new_type = event[u'payload']['ref_type']
event_text.append("{0} deleted {1} <tt>{3}</tt> in {2}."
.format(user_name, new_type, repo_link,
event[u'payload']['ref']))
#DownloadEvent
elif event[u'type'] == "FollowEvent":
color = event_colors['social']
target = event['payload']['target']
try:
event_text.append("{0} is now following {1}."
.format(user_name, target['name']))
except KeyError:
event_text.append("{0} is now following {1}."
.format(user_name, target['login']))
elif event[u'type'] == "ForkEvent":
color = event_colors['branch']
try:
event_text.append("{0} forked {1} to {2}."
.format(user_name, repo_link,
event[u'payload']['forkee']['full_name']))
except KeyError:
event_text.append("{0} forked {1} to {2}/{3}."
.format(user_name, repo_link,
event[u'payload']['forkee']['owner']['login'],
event[u'payload']['forkee']['name']))
event_text.append(repo_desc)
#ForkApplyEvent
elif event[u'type'] == "GistEvent":
event_text.append("{0} {1}d a gist"
.format(user_name, event['payload']['action']))
elif event[u'type'] == "GollumEvent":
event_text.append("{0} updated {2} wiki pages in {1}."
.format(user_name, repo_link,
len(event['payload']['pages'])))
for page in event['payload']['pages']:
event_text.append(page['title'])
elif event[u'type'] == "IssueCommentEvent":
color = event_colors['comment']
issue = Entity.by_name(event['issue'])
event_text.append("{0} commented on issue #{1} in {2}."
.format(user_name, issue['number'], repo_link))
comment = Entity.by_name(event[u'comment'])
event_text.append(issue[u'title'])
event_text.append(comment[u'body'])
elif event[u'type'] == "IssuesEvent":
color = event_colors['issue']
issue = Entity.by_name(event['issue'])
event_text.append("{0} {1} issue #{2} in {3}."
.format(user_name, event['payload']['action'],
issue['number'], repo_link))
event_text.append(issue[u'title'])
elif event[u'type'] == "MemberEvent":
try:
event_text.append("{0} added {1} as a collaborator to {2}."
.format(user_name, event['payload']['member']['name'],
repo_link))
except KeyError:
event_text.append("{0} added {1} as a collaborator to {2}."
.format(user_name, event['payload']['member']['login'],
repo_link))
event_text.append(repo_desc)
elif event[u'type'] == "PublicEvent":
event_text.append("{0} made {1} public."
.format(user_name, repo_link))
event_text.append(repo_desc)
elif event[u'type'] == "PullRequestEvent":
color = event_colors['issue']
# request = Entity.by_name(event['request'])
event_text.append("{0} {1} pull request #{2} in {3}."
.format(user_name, event['payload']['action'],
event['payload']['number'], repo_link))
elif event['type'] == 'PullRequestReviewCommentEvent':
color = event_colors['social']
event_text.append("{0} commented on an issue in {1}."
.format(user_name, repo_link))
comment = Entity.by_name(event['comment'])
event_text.append(comment['body'])
elif event[u'type'] == "PushEvent":
color = event_colors['commit']
event_text.append("{0} pushed {1} commit(s) to {2}."
.format(user_name, event[u'payload']['size'],
repo_link))
for commit in event[u'payload']['commits']:
event_text.append(u'• ' + commit['message'])
#TeamAddEvent
elif event[u'type'] == "WatchEvent":
color = event_colors['social']
event_text.append("{0} is now watching {1}"
.format(user_name, repo_link))
event_text.append(repo_desc)
else:
event_text.append(event['type'])
self.modify_bg(Gtk.StateType.NORMAL, Gdk.color_parse(color))
event_label = mk_label('\n'.join(event_text))
self.box.pack_start(event_label, False, False, 0)
def twill_handler(self, url, links, login_func):
"""
This function uses twill to download the files defined in links
and passing them to the magic handler to be processed.
"""
self.log.debug('twill_handler(%s)' % locals())
parsed_url = urlparse(url)
hostname = parsed_url[1].replace('www.', '')
parent = Entity.by_name(hostname)
if not parent:
parent = Entity(name=hostname)
DBSession.add(parent)
root = Entity.by_name(u'CIVX')
if not root:
root = Entity(name=u'CIVX')
DBSession.add(root)
parent.parent = root
# See if this entity already exists
entity = Entity.by_name(url)
if entity:
self.log.info('Entity(%r) already exists; skipping.' % url)
return
#DBSession.flush()
for category, link_list in links.items():
dest = [self.config['git_dir'], hostname]
if len(links) == 1:
entity = parent
else:
entity = Entity(name=category)
entity[u'url'] = url
entity[u'repo'] = hostname
entity.parent = parent
dest.append(entity.name)
DBSession.add(entity)
DBSession.flush()
dest = os.path.join(*dest)
if not os.path.isdir(dest):
os.makedirs(dest)
b = self.get_browser()
for link in link_list:
# We might have timed out, try to log in again.
login_func()
b.go(link['href'])
# Try to pick out the filename if there is a query
if link['href'].find('=') >= 0:
filename = urlparse(link['href'])[-2].split('=')[1]
else:
filename = link['href'].split('/')[-1]
filename = os.path.join(dest, filename)
save_html(filename)
file_entity = Entity(name=link.contents[0])
file_entity[u'filename'] = filename
file_entity[u'repo'] = hostname
DBSession.add(file_entity)
file_entity.parent = entity
self.log.debug("Created entity %r (parent %r)" % (
file_entity.name, file_entity.parent.name))
magic = self.call_magic_handler(filename, file_entity)
file_entity[u'magic'] = magic
DBSession.flush()
def populate(self, event):
self.event = event
user = Entity.by_name(event[u'actor'])
user_name = user[u'name']
repo = event[u'repo']
if not Entity.by_name(repo):
repo_link = event[u'repo']
repo_desc = ''
else:
repo = Entity.by_name(repo)
repo_link = '<a href="{0}">{1}</a>'.format(repo['url'],
repo['name'])
repo_desc = repo['description']
self.box.pack_start(
url_to_image(user[u'avatar'], user[u'gravatar'], self.scale),
False, False, 10)
event_colors = {
'commit': "#C9FFC1",
'branch': "#C2C9FF",
'issue': "#FFBAF9",
'comment': "#FFDDBD",
'social': "#FFFF80",
}
event_text = []
color = "#FFFFFF"
if event[u'type'] == "CommitCommentEvent":
color = event_colors['comment']
event_text.append("{0} commented on a commit in {1}.".format(
user_name, repo_link))
comment = Entity.by_name(event[u'comment'])
event_text.append(comment[u'body'])
elif event[u'type'] == "CreateEvent":
color = event_colors['branch']
new_type = event[u'payload']['ref_type']
if new_type == 'repository':
event_text.append("{0} created a new {1}, {2}.".format(
user_name, new_type, repo_link))
else:
event_text.append(
"{0} created {1} <tt>{3}</tt> in {2}.".format(
user_name, new_type, repo_link,
event[u'payload']['ref']))
event_text.append(repo_desc)
elif event[u'type'] == "DeleteEvent":
color = event_colors['branch']
new_type = event[u'payload']['ref_type']
event_text.append("{0} deleted {1} <tt>{3}</tt> in {2}.".format(
user_name, new_type, repo_link, event[u'payload']['ref']))
#DownloadEvent
elif event[u'type'] == "FollowEvent":
color = event_colors['social']
target = event['payload']['target']
try:
event_text.append("{0} is now following {1}.".format(
user_name, target['name']))
except KeyError:
event_text.append("{0} is now following {1}.".format(
user_name, target['login']))
elif event[u'type'] == "ForkEvent":
color = event_colors['branch']
try:
event_text.append("{0} forked {1} to {2}.".format(
user_name, repo_link,
event[u'payload']['forkee']['full_name']))
except KeyError:
event_text.append("{0} forked {1} to {2}/{3}.".format(
user_name, repo_link,
event[u'payload']['forkee']['owner']['login'],
event[u'payload']['forkee']['name']))
event_text.append(repo_desc)
#ForkApplyEvent
elif event[u'type'] == "GistEvent":
event_text.append("{0} {1}d a gist".format(
user_name, event['payload']['action']))
elif event[u'type'] == "GollumEvent":
event_text.append("{0} updated {2} wiki pages in {1}.".format(
user_name, repo_link, len(event['payload']['pages'])))
for page in event['payload']['pages']:
event_text.append(page['title'])
elif event[u'type'] == "IssueCommentEvent":
color = event_colors['comment']
issue = Entity.by_name(event['issue'])
event_text.append("{0} commented on issue #{1} in {2}.".format(
user_name, issue['number'], repo_link))
comment = Entity.by_name(event[u'comment'])
event_text.append(issue[u'title'])
event_text.append(comment[u'body'])
elif event[u'type'] == "IssuesEvent":
color = event_colors['issue']
issue = Entity.by_name(event['issue'])
event_text.append("{0} {1} issue #{2} in {3}.".format(
user_name, event['payload']['action'], issue['number'],
repo_link))
event_text.append(issue[u'title'])
elif event[u'type'] == "MemberEvent":
try:
event_text.append(
"{0} added {1} as a collaborator to {2}.".format(
user_name, event['payload']['member']['name'],
repo_link))
except KeyError:
event_text.append(
"{0} added {1} as a collaborator to {2}.".format(
user_name, event['payload']['member']['login'],
repo_link))
event_text.append(repo_desc)
elif event[u'type'] == "PublicEvent":
event_text.append("{0} made {1} public.".format(
user_name, repo_link))
event_text.append(repo_desc)
elif event[u'type'] == "PullRequestEvent":
color = event_colors['issue']
# request = Entity.by_name(event['request'])
event_text.append("{0} {1} pull request #{2} in {3}.".format(
user_name, event['payload']['action'],
event['payload']['number'], repo_link))
elif event['type'] == 'PullRequestReviewCommentEvent':
color = event_colors['social']
event_text.append("{0} commented on an issue in {1}.".format(
user_name, repo_link))
comment = Entity.by_name(event['comment'])
event_text.append(comment['body'])
elif event[u'type'] == "PushEvent":
color = event_colors['commit']
event_text.append("{0} pushed {1} commit(s) to {2}.".format(
user_name, event[u'payload']['size'], repo_link))
for commit in event[u'payload']['commits']:
event_text.append(u'• ' + commit['message'])
#TeamAddEvent
elif event[u'type'] == "WatchEvent":
color = event_colors['social']
event_text.append("{0} is now watching {1}".format(
user_name, repo_link))
event_text.append(repo_desc)
else:
event_text.append(event['type'])
self.modify_bg(Gtk.StateType.NORMAL, Gdk.color_parse(color))
event_label = mk_label('\n'.join(event_text))
self.box.pack_start(event_label, False, False, 0)
def setUp(self):
engine = create_engine('sqlite:///:memory:')
init_model(engine)
metadata.create_all(engine)
PolyScraper().consume('http://www.data.gov/raw/994')
self.entity = Entity.by_name('data.gov')
def consume(self, url):
"""
This method attempts to scrape a URI. First it tries to figure out the
protocol, then tries to pull a hostname out of the url. Then the git
repo is initialized, and we take a close look at the url.
If the hostnme is known to be tricky, it will have a special handler
method written for it and leave from there. Otherwise it goes through
the general path for its protocol, attempting to find useful data.
When everything is done, the entity is updated and messages are sent
out announcing that the scrape is done.
"""
self.log.debug("PolyScraper(%s)" % url)
start = datetime.utcnow()
# Try to pull a protocol off the URI
protocol_end = url.find("://")
protocol = "http"
if not protocol_end == -1:
protocol = url[:protocol_end]
parsed_url = urlparse(url)
hostname = parsed_url[1].replace('www.', '')
# Set a hostname if none is set.
if not hostname:
hostname = u"localhost"
# See if we already know about this URL
entity = Entity.by_name(url)
if entity:
self.log.info('Entity(%r) already exists' % url)
else:
root = Entity.by_name(u'CIVX')
if not root:
root = Entity(name=u'CIVX')
DBSession.add(root)
DBSession.flush()
parent = Entity.by_name(hostname)
if not parent:
parent = Entity(name=hostname)
DBSession.add(parent)
parent.parent = root
self.log.debug("Created entity %r" % parent.name)
entity = Entity(name=url)
DBSession.add(entity)
# hide the exact url entity from our tree
entity.parent = parent
self.log.debug("Created entity %r" % entity.name)
#self.send_message('civx.knowledge.entities.new', {
# 'msg': 'New entity created: %s' % url
# })
DBSession.flush()
# Initialize a git repo for this data source
entity[u'repo'] = hostname
#entity[u'url'] = url
# Initialize the git repository for this domain
#~ self.init_git_repo(repo=hostname)
DBSession.flush()
# Scrape the url (to a certain depth) for data
num_downloads = 0
# Provide a URL handler method that is called with each file pass
# in the soup entity for the link instead, so we can easily look
# around the DOM and pull out titles, etc.
if hostname in self.url_handlers:
#self.url_handlers[hostname](self, soup_link, file_entity)
self.url_handlers[hostname](self, url)
else:
# If we do not specifically handle this file, take a basic approach
# based on the protocol. These could probably also be split off
# into $protocol_handler methods.
self.log.warning('Cannot find %s URL handler' % hostname)
files = []
if protocol == "ftp":
from ftplib import FTP
self.log.debug("FTP support is not implemented yet.")
elif protocol == "file":
search_path = url[protocol_end+3:]
local_files = []
if os.path.isdir(search_path):
# Find all files in directory
for directory in os.walk(search_path):
dirpath = directory[0]
for filename in directory[2]:
local_files.append(os.path.join(dirpath, filename))
else:
local_files.append(search_path)
dest = os.path.join(self.config['git_dir'], hostname)
# FIXME: what about for links to epa.gov from data.gov?
# we probably want our own epa.gov repo namespace to download
# and extract this to
#if not os.path.isdir(dest):
# self.log.debug("mkdir %s" % dest)
# os.makedirs(dest)
# I think this section is deprecated and unnecessary...
#for ext in extensions.split(','):
## if link.endswith('%s' % ext) or '/%s/' % ext in link:
## entity[u'format'] = ext
# if ext not in civx.model.models[Entity]:
# civx.model.models[Entity][ext] = []
for path in local_files:
#raw = self.download_file(link)
#file_name = os.path.basename(link)
#filename = to_unicode(os.path.join(dest, file_name))
#num_downloads += 1
#shutil.copy(raw, filename)
#self.log.debug("Copied %s to %s" % (raw,
# os.path.join(dest, file_name)))
##file_entity = Entity(name=os.path.basename(file_name))
#file_entity = Entity(name=link)
##file_entity[u'url'] = link
#file_entity[u'filename'] = filename
#file_entity[u'repo'] = hostname
#DBSession.add(file_entity)
#file_entity.parent = entity
##file_entity.parent = parent
#self.log.debug("Created entity %r (parent %r)" % (
# file_entity.name, file_entity.parent.name))
file_path = os.path.split(path)[0]
file_name = os.path.split(path)[1]
self.log.info("%s is a local file" % file_name)
files.append((file_path, file_name, path))
#files.append((os.path.dirname(filename), file_name, filename))
else:
# Assume protocol is http
"""
####
f = urllib2.urlopen(url) # XXX: does this load everything into mem?
if f.info().type == 'text/html':
soup = self.get_soup(f.read())
else: # Assume the url is a link to a direct file
# Save the file to disk.
# throw file at magic handlers
####
"""
soup = self.get_soup(url)
for link, soup_link in self.scrape_files_from_url(url, soup_links=True):
parsed_link = urlparse(link)
file_path = '/'.join(parsed_link[2].split('/')[:-1])
file_name = parsed_link[2].split('/')[-1]
files.append((file_path, file_name, link))
for (file_path, file_name, link) in files:
dest = self.config['git_dir'] + hostname + file_path
local = os.path.exists(link)
# See if this file already exists
file_entity = Entity.by_name(link)
#file_entity = Entity.by_name(os.path.basename(file_name))
if file_entity:
self.log.info('Entity(%r) already exists; skipping.' % link)
continue
# FIXME: what about for links to epa.gov from data.gov?
# we probably want our own epa.gov repo namespace to download
# and extract this to
if not os.path.isdir(dest):
os.makedirs(dest)
# I think this section is deprecated and unnecessary...
for ext in extensions.split(','):
# if link.endswith('%s' % ext) or '/%s/' % ext in link:
# entity[u'format'] = ext
if ext not in civx.model.models[Entity]:
civx.model.models[Entity][ext] = []
raw = self.download_file(link)
filename = os.path.join(dest, file_name)
num_downloads += 1
if local:
self.log.debug("Copied %s to %s" % (raw, filename))
shutil.copy(raw, filename)
else:
self.log.debug("Moved %s to %s" % (raw, filename))
shutil.move(raw, filename)
#file_entity = Entity(name=os.path.basename(file_name))
file_entity = Entity(name=link)
#file_entity[u'url'] = link
file_entity[u'filename'] = filename
file_entity[u'repo'] = hostname
DBSession.add(file_entity)
file_entity.parent = entity
#file_entity.parent = parent
self.log.debug("Created entity %r (parent %r)" % (
file_entity.name, file_entity.parent.name))
# Determine the file magic, and call the appropriate handler
file_entity[u'magic'] = self.call_magic_handler(filename, file_entity)
DBSession.flush()
# To do this stuff we'll need to return an entity from the url handler?
#if 'num_files' in entity.facts:
# num_files = int(entity['num_files'])
# print repr(num_files)
#
# if num_files != num_downloads:
# self.log.info('Downloaded %d more files from previous scrape' %
# num_downloads - num_files)
# entity[u'num_files'] += num_downloads
# else:
#entity[u'num_files'] = num_downloads
#if u'date_added' not in entity.facts:
# entity[u'date_added'] = unicode(datetime.utcnow())
#entity[u'date_last_scraped'] = unicode(datetime.utcnow())
if 'changelog' not in entity.facts:
entity[u'changelog'] = []
finish = datetime.utcnow()
changelog = {
u'start_time': unicode(start),
u'finish_time': unicode(finish),
u'elapsed_time': unicode(finish-start),
u'num_downloads': num_downloads,
#u'num_children': len(entity.children),
#~ u'git_commit': self.get_latest_commit_id(),
}
entity[u'changelog'].append(changelog)
DBSession.commit()
self.log.info("== Statistics ==")
self.log.info("Scraped url: " + url)
self.log.info("Number of downloaded files: %d" % num_downloads)
def test_associating_facts_unicode_by_get(self):
apple = Entity('apple')
DBSession.add(apple)
DBSession.commit()
apple['foo'] = u'bar'
eq_(apple.get('foo'), 'bar')
def test_associating_facts_get_custom_default(self):
apple = Entity('apple')
DBSession.add(apple)
DBSession.commit()
apple['foo'] = u'bar'
eq_(apple.get('baz', 'zomg'), 'zomg')
def data_gov_handler(self, url):
""" data.gov handler.
Entity(CIVX)
|-Entity(data.gov)
| |-Entity(Agency)
| |-Entity(http://www.data.gov/raw/674)
| | |-Fact(title), ...
| | |-Entity(http://www.epa.gov/tri/tridata/tri08/early_data/statedata/basic/TRI_2008_TN_v08.exe)
"""
self.log.debug('data_gov_handler(%s)' % locals())
parsed_url = urlparse(url)
hostname = parsed_url[1].replace('www.', '')
data_types = ('csv', 'RDF', 'xml', 'kml', 'PDF', 'shapefile', 'XLS')
fields = ('Agency', 'Sub-Agency', 'Category', 'Date Released',
'Date Updated', 'Time Period', 'Frequency',
'Description', 'Data.gov Data Category Type',
'Specialized Data Category Designation',
'Keywords', 'Unique ID', 'Citation',
'Agency Program Page', 'Agency Data Series Page',
'Unit of Analysis', 'Granularity', 'Geographic Coverage',
'Collection Mode', 'Data Collection Instrument',
'Data Dictionary/Variable List', 'Technical Documentation',
'Additional Metadata')
# Our top-level data.gov entity
data_gov = Entity.by_name('data.gov')
if not data_gov:
data_gov = Entity(name=u'data.gov')
DBSession.add(data_gov)
root = Entity.by_name(u'CIVX')
if not root:
root = Entity(name=u'CIVX')
DBSession.add(root)
data_gov.parent = root
# See if this entity already exists
#~ entity = Entity.by_name(url)
#~ if entity:
#~ self.log.info('Entity(%r) already exists; skipping.' % url)
#~ return
soup = self.get_soup(url)
# If this is a raw data profile, grab the title of the dataset
if '/raw/' in url:
# Create a new Entity for this URL
title = soup.find('h2', {'id': 'datasetName'}).string.decode('utf-8', 'replace')
entity = Entity(name=title)
entity[u'url'] = url
entity[u'repo'] = hostname
DBSession.add(entity)
dest = [self.config['git_dir'], hostname]
# Extract data for each field
for field in fields:
data = soup.find(text=field)
if data and data.next and data.next.next:
data = data.next.next.string
if data:
entity[unicode(field)] = data.decode('utf-8').strip()
DBSession.flush()
# Create seperate parent Agency Entity
if u'Agency' in entity.facts:
agency = Entity(name=entity[u'Agency'])
agency.parent = data_gov
parent = agency
DBSession.add(agency)
dest.append(entity[u'Agency'])
if u'Sub-Agency' in entity.facts:
subagency = Entity(name=entity[u'Sub-Agency'])
subagency.parent = agency
parent = subagency
DBSession.add(subagency)
dest.append(entity[u'Sub-Agency'])
DBSession.flush()
# Have the URL be the child of the agency or sub-agency
entity.parent = parent
# Elegant repo paths: data.gov/Agency[/Sub-Agency]/title/filename
dest.append(entity.name)
dest = os.path.join(*dest)
if not os.path.isdir(dest):
os.makedirs(dest)
# Scrape all available raw data types
downloads = soup.find_all('a', href=re.compile(r'^/download'))
for button in downloads:
data = button.string.split()[0]
link = button['href']
if link:
link = urljoin('http://explore.data.gov', link)
entity[data.lower()] = link
parsed_link = urlparse(link)
file_name = parsed_link[2].split('/')[-1]
raw = self.download_file(link)
filename = os.path.join(dest, file_name)
shutil.move(raw, filename)
self.log.debug("Moved %s to %s" % (raw, filename))
# Create a new entity for this file
file_entity = Entity(name=link)
DBSession.add(file_entity)
file_entity[u'filename'] = filename
file_entity.parent = entity
# Process this file accordingly
self.call_magic_handler(filename, file_entity)
# Find external map links
map = soup.find('a', href=re.compile(r'^/externallink/map/'))
if map:
map = urllib.unquote(map.get('href', '')[18:]).split('/')[0].replace('###', '/')
entity[u'map'] = map
DBSession.flush()
# If this is from a table of results, grab the title fom this row
else:
self.log.debug("entity[url] = %r" % entity[u'url'])
raise NotImplementedError("Scraping titles from data.gov tables not yet supported")
