def import_speeches(abbr, data_dir):
data_dir = os.path.join(data_dir, abbr)
pattern = os.path.join(data_dir, 'speeches', '*.json')
speech_record_ids = defaultdict(set)
for path in glob.iglob(pattern):
# OK, We need to first go through all the JSON and load the document
# IDs to clear out.
with open(path) as f:
data = prepare_obj(json.load(f))
session = data['session']
chamber = data['chamber']
speech_record_ids[session].add((chamber, data['record_id']))
for session in speech_record_ids:
for obj in speech_record_ids[session]:
chamber, record = obj
# XXX: Should we really be clearing them all up front? Should
# we clear as we process each record block? Is it OK to
# store everything in memory? (there's a lot)
#
# this will result in broken data if the import breaks
# below.
# -- PRT
clear_old_speeches(abbr, chamber, session, record)
for path in glob.iglob(pattern):
# OK, now we need to import all the JSON. We don't keep the objects
# from above, since that'd really dent memory, and a few more ms on
# import isn't the end of the world.
with open(path) as f:
data = prepare_obj(json.load(f))
import_speech(data)
def import_speeches(abbr, data_dir):
data_dir = os.path.join(data_dir, abbr)
pattern = os.path.join(data_dir, 'speeches', '*.json')
speech_record_ids = defaultdict(set)
for path in glob.iglob(pattern):
# OK, We need to first go through all the JSON and load the document
# IDs to clear out.
with open(path) as f:
data = prepare_obj(json.load(f))
session = data['session']
chamber = data['chamber']
speech_record_ids[session].add((chamber, data['record_id']))
for session in speech_record_ids:
for obj in speech_record_ids[session]:
chamber, record = obj
# XXX: Should we really be clearing them all up front? Should
# we clear as we process each record block? Is it OK to
# store everything in memory? (there's a lot)
#
# this will result in broken data if the import breaks
# below.
# -- PRT
clear_old_speeches(abbr, chamber, session, record)
for path in glob.iglob(pattern):
# OK, now we need to import all the JSON. We don't keep the objects
# from above, since that'd really dent memory, and a few more ms on
# import isn't the end of the world.
with open(path) as f:
data = prepare_obj(json.load(f))
import_speech(data)
def import_events(state, data_dir, import_actions=True):
data_dir = os.path.join(data_dir, state)
pattern = os.path.join(data_dir, 'events', '*.json')
for path in glob.iglob(pattern):
with open(path) as f:
data = prepare_obj(json.load(f))
event = None
if '_guid' in data:
event = db.events.find_one({'state': data['state'],
'_guid': data['_guid']})
if not event:
event = db.events.find_one({'state': data['state'],
'when': data['when'],
'end': data['end'],
'type': data['type'],
'description': data['description']})
if not event:
data['created_at'] = datetime.datetime.utcnow()
data['updated_at'] = data['created_at']
_insert_with_id(data)
else:
update(event, data, db.events)
# if import_actions:
# actions_to_events(state)
ensure_indexes()
def import_bills(abbr, data_dir):
data_dir = os.path.join(data_dir, abbr)
pattern = os.path.join(data_dir, 'bills', '*.json')
git_prelod(abbr)
counts = {"update": 0, "insert": 0, "total": 0}
votes = load_standalone_votes(data_dir)
try:
categorizer = SubjectCategorizer(abbr)
except Exception as e:
logger.debug('Proceeding without subject categorizer: %s' % e)
categorizer = None
paths = glob.glob(pattern)
for path in paths:
with open(path) as f:
data = prepare_obj(json.load(f))
counts["total"] += 1
ret = import_bill(data, votes, categorizer)
counts[ret] += 1
logger.info('imported %s bill files' % len(paths))
for remaining in votes.keys():
logger.debug('Failed to match vote %s %s %s' %
tuple([r.encode('ascii', 'replace') for r in remaining]))
populate_current_fields(abbr)
git_commit("Import Update")
return counts
def import_committees(abbr, data_dir):
data_dir = os.path.join(data_dir, abbr)
pattern = os.path.join(data_dir, 'committees', '*.json')
counts = {"update": 0, "insert": 0, "total": 0}
meta = db.metadata.find_one({'_id': abbr})
current_term = meta['terms'][-1]['name']
current_session = meta['terms'][-1]['sessions'][-1]
paths = glob.glob(pattern)
for committee in db.committees.find({settings.LEVEL_FIELD: abbr}):
committee['members'] = []
db.committees.save(committee, safe=True)
# import committees from legislator roles, no standalone committees scraped
if not paths:
import_committees_from_legislators(current_term, abbr)
for path in paths:
with open(path) as f:
data = prepare_obj(json.load(f))
counts["total"] += 1
ret = import_committee(data, current_session, current_term)
counts[ret] += 1
logger.info('imported %s committee files' % len(paths))
link_parents(abbr)
ensure_indexes()
return counts
def import_committees(abbr, data_dir):
data_dir = os.path.join(data_dir, abbr)
pattern = os.path.join(data_dir, 'committees', '*.json')
meta = db.metadata.find_one({'_id': abbr})
current_term = meta['terms'][-1]['name']
current_session = meta['terms'][-1]['sessions'][-1]
level = meta['level']
paths = glob.glob(pattern)
for committee in db.committees.find({'level': level, level: abbr}):
committee['members'] = []
db.committees.save(committee, safe=True)
# import committees from legislator roles, no standalone committees scraped
if not paths:
import_committees_from_legislators(current_term, level, abbr)
for path in paths:
with open(path) as f:
data = prepare_obj(json.load(f))
import_committee(data, current_session, current_term)
print 'imported %s committee files' % len(paths)
link_parents(level, abbr)
ensure_indexes()
def import_bills(abbr, data_dir):
data_dir = os.path.join(data_dir, abbr)
pattern = os.path.join(data_dir, "bills", "*.json")
counts = {"update": 0, "insert": 0, "total": 0}
votes = import_votes(data_dir)
try:
categorizer = SubjectCategorizer(abbr)
except Exception as e:
logger.debug("Proceeding without subject categorizer: %s" % e)
categorizer = None
paths = glob.glob(pattern)
for path in paths:
with open(path) as f:
data = prepare_obj(json.load(f))
counts["total"] += 1
ret = import_bill(data, votes, categorizer)
counts[ret] += 1
logger.info("imported %s bill files" % len(paths))
for remaining in votes.keys():
logger.debug("Failed to match vote %s %s %s" % tuple([r.encode("ascii", "replace") for r in remaining]))
meta = db.metadata.find_one({"_id": abbr})
level = meta["level"]
populate_current_fields(level, abbr)
ensure_indexes()
return counts
def import_bills(abbr, data_dir):
data_dir = os.path.join(data_dir, abbr)
pattern = os.path.join(data_dir, 'bills', '*.json')
counts = {"update": 0, "insert": 0, "total": 0}
votes = import_votes(data_dir)
try:
categorizer = SubjectCategorizer(abbr)
except Exception as e:
logger.debug('Proceeding without subject categorizer: %s' % e)
categorizer = None
paths = glob.glob(pattern)
for path in paths:
with open(path) as f:
data = prepare_obj(json.load(f))
counts["total"] += 1
ret = import_bill(data, votes, categorizer)
counts[ret] += 1
logger.info('imported %s bill files' % len(paths))
for remaining in votes.keys():
logger.debug('Failed to match vote %s %s %s' %
tuple([r.encode('ascii', 'replace') for r in remaining]))
meta = db.metadata.find_one({'_id': abbr})
level = meta['level']
populate_current_fields(level, abbr)
ensure_indexes()
return counts
def import_events(abbr, data_dir, import_actions=False):
data_dir = os.path.join(data_dir, abbr)
pattern = os.path.join(data_dir, 'events', '*.json')
for path in glob.iglob(pattern):
with open(path) as f:
data = prepare_obj(json.load(f))
def _resolve_ctty(committee):
return get_committee_id(data[settings.LEVEL_FIELD],
committee['chamber'],
committee['participant'])
def _resolve_leg(leg):
chamber = leg['chamber'] if leg['chamber'] in ['upper', 'lower'] \
else None
return get_legislator_id(abbr, data['session'], chamber,
leg['participant'])
resolvers = {"committee": _resolve_ctty, "legislator": _resolve_leg}
for entity in data['participants']:
type = entity['participant_type']
id = None
if type in resolvers:
id = resolvers[type](entity)
else:
logger.warning("I don't know how to resolve a %s" % type)
entity['id'] = id
for bill in data['related_bills']:
bill['_scraped_bill_id'] = bill['bill_id']
bill_id = bill['bill_id']
bill_id = fix_bill_id(bill_id)
bill['bill_id'] = ""
db_bill = db.bills.find_one({
"$or": [{
settings.LEVEL_FIELD: abbr,
'session': data['session'],
'bill_id': bill_id
}, {
settings.LEVEL_FIELD: abbr,
'session': data['session'],
'alternate_bill_ids': bill_id
}]
})
if not db_bill:
logger.warning("Error: Can't find %s" % bill_id)
db_bill = {}
db_bill['_id'] = None
# Events are really hard to pin to a chamber. Some of these are
# also a committee considering a bill from the other chamber, or
# something like that.
bill['bill_id'] = db_bill['_id']
import_event(data)
ensure_indexes()
def import_legislator(data):
data = prepare_obj(data)
data['_scraped_name'] = data['full_name']
# Rename 'role' -> 'type'
for role in data['roles']:
if 'role' in role:
role['type'] = role.pop('role')
# copy over country and/or state into role
# TODO: base this on all possible level fields
role['level'] = data['level']
if 'country' in data:
role['country'] = data['country']
if 'state' in data:
role['state'] = data['state']
cur_role = data['roles'][0]
term = cur_role['term']
level = data['level']
abbrev = data[level]
prev_term = get_previous_term(abbrev, term)
next_term = get_next_term(abbrev, term)
spec = {level: abbrev,
'type': cur_role['type'],
'term': {'$in': [term, prev_term, next_term]}}
if 'district' in cur_role:
spec['district'] = cur_role['district']
if 'chamber' in cur_role:
spec['chamber'] = cur_role['chamber']
leg = db.legislators.find_one(
{'level': level, level: abbrev,
'_scraped_name': data['full_name'],
'roles': {'$elemMatch': spec}})
if leg:
if 'old_roles' not in leg:
leg['old_roles'] = {}
if leg['roles'][0]['term'] == prev_term:
# Move to old
leg['old_roles'][leg['roles'][0]['term']] = leg['roles']
elif leg['roles'][0]['term'] == next_term:
leg['old_roles'][term] = data['roles']
data['roles'] = leg['roles']
update(leg, data, db.legislators)
return "update"
else:
insert_with_id(data)
return "insert"
def import_events(abbr, data_dir, import_actions=False):
data_dir = os.path.join(data_dir, abbr)
pattern = os.path.join(data_dir, 'events', '*.json')
for path in glob.iglob(pattern):
with open(path) as f:
data = prepare_obj(json.load(f))
import_event(data)
ensure_indexes()
def import_events(abbr, data_dir, import_actions=False):
data_dir = os.path.join(data_dir, abbr)
pattern = os.path.join(data_dir, 'events', '*.json')
for path in glob.iglob(pattern):
with open(path) as f:
data = prepare_obj(json.load(f))
import_event(data)
ensure_indexes()
def import_legislator(data):
data = prepare_obj(data)
data['_scraped_name'] = data['full_name']
# Rename 'role' -> 'type'
for role in data['roles']:
if 'role' in role:
role['type'] = role['role']
del role['role']
# copy over country and/or state into role
# TODO: base this on all possible level fields
role['level'] = data['level']
if 'country' in data:
role['country'] = data['country']
if 'state' in data:
role['state'] = data['state']
cur_role = data['roles'][0]
term = cur_role['term']
level = data['level']
abbrev = data[level]
prev_term = get_previous_term(abbrev, term)
next_term = get_next_term(abbrev, term)
spec = {level: abbrev,
'type': cur_role['type'],
'term': {'$in': [term, prev_term, next_term]}}
if 'district' in cur_role:
spec['district'] = cur_role['district']
if 'chamber' in cur_role:
spec['chamber'] = cur_role['chamber']
leg = db.legislators.find_one(
{'level': level, level: abbrev,
'_scraped_name': data['full_name'],
'roles': {'$elemMatch': spec}})
if leg:
if 'old_roles' not in leg:
leg['old_roles'] = {}
if leg['roles'][0]['term'] == prev_term:
# Move to old
leg['old_roles'][leg['roles'][0]['term']] = leg['roles']
elif leg['roles'][0]['term'] == next_term:
leg['old_roles'][term] = data['roles']
data['roles'] = leg['roles']
update(leg, data, db.legislators)
else:
insert_with_id(data)
def import_events(abbr, data_dir, import_actions=False):
data_dir = os.path.join(data_dir, abbr)
pattern = os.path.join(data_dir, "events", "*.json")
for path in glob.iglob(pattern):
with open(path) as f:
data = prepare_obj(json.load(f))
def _resolve_ctty(committee):
return get_committee_id(data[settings.LEVEL_FIELD], committee["chamber"], committee["participant"])
def _resolve_leg(leg):
chamber = leg["chamber"] if leg["chamber"] in ["upper", "lower"] else None
return get_legislator_id(abbr, data["session"], chamber, leg["participant"])
resolvers = {"committee": _resolve_ctty, "legislator": _resolve_leg}
for entity in data["participants"]:
type = entity["participant_type"]
id = None
if type in resolvers:
id = resolvers[type](entity)
else:
logger.warning("I don't know how to resolve a %s" % type)
entity["id"] = id
for bill in data["related_bills"]:
bill["_scraped_bill_id"] = bill["bill_id"]
bill_id = bill["bill_id"]
bill_id = fix_bill_id(bill_id)
bill["bill_id"] = ""
db_bill = db.bills.find_one(
{
"$or": [
{settings.LEVEL_FIELD: abbr, "session": data["session"], "bill_id": bill_id},
{settings.LEVEL_FIELD: abbr, "session": data["session"], "alternate_bill_ids": bill_id},
]
}
)
if not db_bill:
logger.warning("Error: Can't find %s" % bill_id)
db_bill = {}
db_bill["_id"] = None
# Events are really hard to pin to a chamber. Some of these are
# also a committee considering a bill from the other chamber, or
# something like that.
bill["bill_id"] = db_bill["_id"]
import_event(data)
ensure_indexes()
def import_legislator(data):
data = prepare_obj(data)
data['_scraped_name'] = data['full_name']
# Rename 'role' -> 'type'
for role in data['roles']:
if 'role' in role:
role['type'] = role['role']
del role['role']
cur_role = data['roles'][0]
term = cur_role['term']
prev_term = get_previous_term(data['state'], term)
next_term = get_next_term(data['state'], term)
spec = {
'state': data['state'],
'type': cur_role['type'],
'term': {
'$in': [term, prev_term, next_term]
}
}
if 'district' in cur_role:
spec['district'] = cur_role['district']
if 'chamber' in cur_role:
spec['chamber'] = cur_role['chamber']
leg = db.legislators.find_one({
'state': data['state'],
'_scraped_name': data['full_name'],
'roles': {
'$elemMatch': spec
}
})
if leg:
if 'old_roles' not in leg:
leg['old_roles'] = {}
if leg['roles'][0]['term'] == prev_term:
# Move to old
leg['old_roles'][leg['roles'][0]['term']] = leg['roles']
elif leg['roles'][0]['term'] == next_term:
leg['old_roles'][term] = data['roles']
data['roles'] = leg['roles']
update(leg, data, db.legislators)
else:
insert_with_id(data)
def import_votes(state, data_dir):
data_dir = os.path.join(data_dir, state)
pattern = os.path.join(data_dir, 'votes', '*.json')
paths = glob.glob(pattern)
for path in paths:
with open(path) as f:
data = prepare_obj(json.load(f))
# clean up bill_id, needs to match the one already in the database
data['bill_id'] = fix_bill_id(data['bill_id'])
bill = db.bills.find_one({'state': state,
'chamber': data['bill_chamber'],
'session': data['session'],
'bill_id': data['bill_id']})
if not bill:
_log.warning("Couldn't find bill %s" % data['bill_id'])
continue
del data['bill_id']
try:
del data['filename']
except KeyError:
pass
for vtype in ('yes_votes', 'no_votes', 'other_votes'):
svlist = []
for svote in data[vtype]:
id = get_legislator_id(state, data['session'],
data['chamber'], svote)
svlist.append({'name': svote, 'leg_id': id})
data[vtype] = svlist
for vote in bill['votes']:
if (vote['motion'] == data['motion']
and vote['date'] == data['date']):
vote.update(data)
break
else:
bill['votes'].append(data)
db.bills.save(bill, safe=True)
print 'imported %s vote files' % len(paths)
def import_metadata(state, data_dir):
preserved = {}
old_metadata = db.metadata.find_one({'_id':state}) or {}
for field in PRESERVED_FIELDS:
if field in old_metadata:
preserved[field] = old_metadata[field]
data_dir = os.path.join(data_dir, state)
with open(os.path.join(data_dir, 'state_metadata.json')) as f:
data = json.load(f)
data['_type'] = 'metadata'
data = prepare_obj(data)
data['_id'] = state
data.update(preserved)
db.metadata.save(data, safe=True)
def import_votes(data_dir):
pattern = os.path.join(data_dir, "votes", "*.json")
paths = glob.glob(pattern)
votes = defaultdict(list)
for path in paths:
with open(path) as f:
data = prepare_obj(json.load(f))
# need to match bill_id already in the database
bill_id = fix_bill_id(data.pop("bill_id"))
votes[(data["bill_chamber"], data["session"], bill_id)].append(data)
logger.info("imported %s vote files" % len(paths))
return votes
def load_standalone_votes(data_dir):
pattern = os.path.join(data_dir, 'votes', '*.json')
paths = glob.glob(pattern)
votes = defaultdict(list)
for path in paths:
with open(path) as f:
data = prepare_obj(json.load(f))
# need to match bill_id already in the database
bill_id = fix_bill_id(data.pop('bill_id'))
votes[(data['bill_chamber'], data['session'], bill_id)].append(data)
logger.info('imported %s vote files' % len(paths))
return votes
def import_votes(state, data_dir):
pattern = os.path.join(data_dir, 'votes', '*.json')
paths = glob.glob(pattern)
votes = defaultdict(list)
for path in paths:
with open(path) as f:
data = prepare_obj(json.load(f))
# need to match bill_id already in the database
bill_id = fix_bill_id(data.pop('bill_id'))
votes[(data['bill_chamber'], data['session'], bill_id)].append(data)
print 'imported %s vote files' % len(paths)
return votes
def import_events(abbr, data_dir, import_actions=False):
data_dir = os.path.join(data_dir, abbr)
pattern = os.path.join(data_dir, 'events', '*.json')
for path in glob.iglob(pattern):
with open(path) as f:
data = prepare_obj(json.load(f))
for committee in data['participants']:
cttyid = get_committee_id(data['level'], data['state'],
committee['participant'],
committee['chamber'] )
if cttyid:
committee['committee_id'] = cttyid
for bill in data['related_bills']:
bill['_scraped_bill_id'] = bill['bill_id']
bill_id = bill['bill_id']
bill_id = fix_bill_id(bill_id)
bill['bill_id'] = ""
db_bill = db.bills.find_one({
"$or": [
{
"state": abbr,
'session': data['session'],
'bill_id': bill_id
},
{
"state": abbr,
'session': data['session'],
'alternate_bill_ids': bill_id
}
]
})
if not db_bill:
logger.warning("Error: Can't find %s" % bill_id)
db_bill = {}
db_bill['_id'] = None
# Events are really hard to pin to a chamber. Some of these are
# also a committee considering a bill from the other chamber, or
# something like that.
bill['bill_id'] = db_bill['_id']
import_event(data)
ensure_indexes()
def import_metadata(state, data_dir):
preserved = {}
old_metadata = db.metadata.find_one({'_id':state}) or {}
for field in PRESERVED_FIELDS:
if field in old_metadata:
preserved[field] = old_metadata[field]
data_dir = os.path.join(data_dir, state)
with open(os.path.join(data_dir, 'state_metadata.json')) as f:
data = json.load(f)
data['_type'] = 'metadata'
data = prepare_obj(data)
data['_id'] = state
data.update(preserved)
data['latest_update'] = datetime.datetime.utcnow()
db.metadata.save(data, safe=True)
def import_metadata(abbr, data_dir):
preserved = {}
old_metadata = db.metadata.find_one({'_id': abbr}) or {}
for field in PRESERVED_FIELDS:
if field in old_metadata:
preserved[field] = old_metadata[field]
data_dir = os.path.join(data_dir, abbr)
with open(os.path.join(data_dir, 'metadata.json')) as f:
data = json.load(f)
data['_type'] = 'metadata'
data = prepare_obj(data)
data['_id'] = abbr
data.update(preserved)
data['latest_update'] = datetime.datetime.utcnow()
db.metadata.save(data, safe=True)
def import_legislator(data):
data = prepare_obj(data)
data['_scraped_name'] = data['full_name']
# Rename 'role' -> 'type'
for role in data['roles']:
if 'role' in role:
role['type'] = role['role']
del role['role']
cur_role = data['roles'][0]
term = cur_role['term']
prev_term = get_previous_term(data['state'], term)
next_term = get_next_term(data['state'], term)
spec = {'state': data['state'],
'type': cur_role['type'],
'term': {'$in': [term, prev_term, next_term]}}
if 'district' in cur_role:
spec['district'] = cur_role['district']
if 'chamber' in cur_role:
spec['chamber'] = cur_role['chamber']
leg = db.legislators.find_one(
{'state': data['state'],
'_scraped_name': data['full_name'],
'roles': {'$elemMatch': spec}})
if leg:
if 'old_roles' not in leg:
leg['old_roles'] = {}
if leg['roles'][0]['term'] == prev_term:
# Move to old
leg['old_roles'][leg['roles'][0]['term']] = leg['roles']
elif leg['roles'][0]['term'] == next_term:
leg['old_roles'][term] = data['roles']
data['roles'] = leg['roles']
update(leg, data, db.legislators)
else:
insert_with_id(data)
def import_bills(abbr, data_dir):
data_dir = os.path.join(data_dir, abbr)
pattern = os.path.join(data_dir, 'bills', '*.json')
git_prelod(abbr)
counts = {
"update": 0,
"insert": 0,
"total": 0
}
votes = load_standalone_votes(data_dir)
try:
categorizer = SubjectCategorizer(abbr)
except Exception as e:
logger.debug('Proceeding without subject categorizer: %s' % e)
categorizer = None
paths = glob.glob(pattern)
for path in paths:
with open(path) as f:
data = prepare_obj(json.load(f))
counts["total"] += 1
ret = import_bill(data, votes, categorizer)
counts[ret] += 1
logger.info('imported %s bill files' % len(paths))
for remaining in votes.keys():
logger.debug('Failed to match vote %s %s %s' % tuple([
r.encode('ascii', 'replace') for r in remaining]))
populate_current_fields(abbr)
git_commit("Import Update")
ensure_indexes()
return counts
def import_committees(abbr, data_dir):
data_dir = os.path.join(data_dir, abbr)
pattern = os.path.join(data_dir, 'committees', '*.json')
counts = {
"update": 0,
"insert": 0,
"total": 0
}
meta = db.metadata.find_one({'_id': abbr})
current_term = meta['terms'][-1]['name']
current_session = meta['terms'][-1]['sessions'][-1]
paths = glob.glob(pattern)
for committee in db.committees.find({settings.LEVEL_FIELD: abbr}):
committee['members'] = []
db.committees.save(committee, safe=True)
# import committees from legislator roles, no standalone committees scraped
if not paths:
import_committees_from_legislators(current_term, abbr)
for path in paths:
with open(path) as f:
data = prepare_obj(json.load(f))
counts["total"] += 1
ret = import_committee(data, current_session, current_term)
counts[ret] += 1
logger.info('imported %s committee files' % len(paths))
link_parents(abbr)
ensure_indexes()
return counts
def import_bills(abbr, data_dir):
data_dir = os.path.join(data_dir, abbr)
pattern = os.path.join(data_dir, 'bills', '*.json')
votes = import_votes(data_dir)
paths = glob.glob(pattern)
for path in paths:
with open(path) as f:
data = prepare_obj(json.load(f))
import_bill(data, votes)
print 'imported %s bill files' % len(paths)
for remaining in votes.keys():
print 'Failed to match vote %s %s %s' % tuple([
r.encode('ascii', 'replace') for r in remaining])
meta = db.metadata.find_one({'_id': abbr})
level = meta['level']
populate_current_fields(level, abbr)
ensure_indexes()
def import_bills(abbr, data_dir):
data_dir = os.path.join(data_dir, abbr)
pattern = os.path.join(data_dir, 'bills', '*.json')
votes = import_votes(data_dir)
paths = glob.glob(pattern)
for path in paths:
with open(path) as f:
data = prepare_obj(json.load(f))
import_bill(data, votes)
print 'imported %s bill files' % len(paths)
for remaining in votes.keys():
print 'Failed to match vote %s %s %s' % tuple(
[r.encode('ascii', 'replace') for r in remaining])
meta = db.metadata.find_one({'_id': abbr})
level = meta['level']
populate_current_fields(level, abbr)
ensure_indexes()
def import_bills(state, data_dir):
data_dir = os.path.join(data_dir, state)
pattern = os.path.join(data_dir, "bills", "*.json")
meta = db.metadata.find_one({"_id": state})
# Build a session to term mapping
sessions = {}
for term in meta["terms"]:
for session in term["sessions"]:
sessions[session] = term["name"]
votes = import_votes(state, data_dir)
paths = glob.glob(pattern)
for path in paths:
with open(path) as f:
data = prepare_obj(json.load(f))
# clean up bill_id
data["bill_id"] = fix_bill_id(data["bill_id"])
# move subjects to scraped_subjects
subjects = data.pop("subjects", None)
# NOTE: intentionally doesn't copy blank lists of subjects
# this avoids the problem where a bill is re-run but we can't
# get subjects anymore (quite common in fact)
if subjects:
data["scraped_subjects"] = subjects
# add loaded votes to data
bill_votes = votes.pop((data["chamber"], data["session"], data["bill_id"]), [])
data["votes"].extend(bill_votes)
bill = db.bills.find_one(
{"state": data["state"], "session": data["session"], "chamber": data["chamber"], "bill_id": data["bill_id"]}
)
vote_matcher = VoteMatcher(data["state"])
if bill:
vote_matcher.learn_vote_ids(bill["votes"])
vote_matcher.set_vote_ids(data["votes"])
# match sponsor leg_ids
for sponsor in data["sponsors"]:
id = get_legislator_id(state, data["session"], None, sponsor["name"])
sponsor["leg_id"] = id
for vote in data["votes"]:
# committee_ids
if "committee" in vote:
committee_id = get_committee_id(state, vote["chamber"], vote["committee"])
vote["committee_id"] = committee_id
# vote leg_ids
for vtype in ("yes_votes", "no_votes", "other_votes"):
svlist = []
for svote in vote[vtype]:
id = get_legislator_id(state, data["session"], vote["chamber"], svote)
svlist.append({"name": svote, "leg_id": id})
vote[vtype] = svlist
data["_term"] = sessions[data["session"]]
# Merge any version titles into the alternate_titles list
alt_titles = set(data.get("alternate_titles", []))
for version in data["versions"]:
if "title" in version:
alt_titles.add(version["title"])
if "+short_title" in version:
alt_titles.add(version["+short_title"])
try:
# Make sure the primary title isn't included in the
# alternate title list
alt_titles.remove(data["title"])
except KeyError:
pass
data["alternate_titles"] = list(alt_titles)
if not bill:
data["_keywords"] = list(bill_keywords(data))
insert_with_id(data)
else:
data["_keywords"] = list(bill_keywords(data))
update(bill, data, db.bills)
print "imported %s bill files" % len(paths)
for remaining in votes.keys():
print "Failed to match vote %s %s %s" % tuple([r.encode("ascii", "replace") for r in remaining])
populate_current_fields(state)
ensure_indexes()
def import_bills(state, data_dir):
data_dir = os.path.join(data_dir, state)
pattern = os.path.join(data_dir, 'bills', '*.json')
meta = db.metadata.find_one({'_id': state})
# Build a session to term mapping
sessions = {}
for term in meta['terms']:
for session in term['sessions']:
sessions[session] = term['name']
votes = import_votes(state, data_dir)
paths = glob.glob(pattern)
for path in paths:
with open(path) as f:
data = prepare_obj(json.load(f))
# clean up bill_id
data['bill_id'] = fix_bill_id(data['bill_id'])
# move subjects to scraped_subjects
subjects = data.pop('subjects', None)
if subjects:
data['scraped_subjects'] = subjects
# add loaded votes to data
bill_votes = votes.pop(
(data['chamber'], data['session'], data['bill_id']), [])
data['votes'].extend(bill_votes)
bill = db.bills.find_one({
'state': data['state'],
'session': data['session'],
'chamber': data['chamber'],
'bill_id': data['bill_id']
})
vote_matcher = VoteMatcher(data['state'])
if bill:
vote_matcher.learn_vote_ids(bill['votes'])
vote_matcher.set_vote_ids(data['votes'])
# match sponsor leg_ids
for sponsor in data['sponsors']:
id = get_legislator_id(state, data['session'], None,
sponsor['name'])
sponsor['leg_id'] = id
for vote in data['votes']:
# committee_ids
if 'committee' in vote:
committee_id = get_committee_id(state, vote['chamber'],
vote['committee'])
vote['committee_id'] = committee_id
# vote leg_ids
for vtype in ('yes_votes', 'no_votes', 'other_votes'):
svlist = []
for svote in vote[vtype]:
id = get_legislator_id(state, data['session'],
vote['chamber'], svote)
svlist.append({'name': svote, 'leg_id': id})
vote[vtype] = svlist
data['_term'] = sessions[data['session']]
# Merge any version titles into the alternate_titles list
alt_titles = set(data.get('alternate_titles', []))
for version in data['versions']:
if 'title' in version:
alt_titles.add(version['title'])
if '+short_title' in version:
alt_titles.add(version['+short_title'])
try:
# Make sure the primary title isn't included in the
# alternate title list
alt_titles.remove(data['title'])
except KeyError:
pass
data['alternate_titles'] = list(alt_titles)
if not bill:
data['_keywords'] = list(bill_keywords(data))
insert_with_id(data)
else:
data['_keywords'] = list(bill_keywords(data))
update(bill, data, db.bills)
print 'imported %s bill files' % len(paths)
for remaining in votes.keys():
print 'Failed to match vote %s %s %s' % tuple(
[r.encode('ascii', 'replace') for r in remaining])
populate_current_fields(state)
ensure_indexes()
def import_bills(state, data_dir):
data_dir = os.path.join(data_dir, state)
pattern = os.path.join(data_dir, 'bills', '*.json')
meta = db.metadata.find_one({'_id': state})
# Build a session to term mapping
sessions = {}
for term in meta['terms']:
for session in term['sessions']:
sessions[session] = term['name']
votes = import_votes(state, data_dir)
paths = glob.glob(pattern)
for path in paths:
with open(path) as f:
data = prepare_obj(json.load(f))
# clean up bill_id
data['bill_id'] = fix_bill_id(data['bill_id'])
# move subjects to scraped_subjects
subjects = data.pop('subjects', None)
# NOTE: intentionally doesn't copy blank lists of subjects
# this avoids the problem where a bill is re-run but we can't
# get subjects anymore (quite common in fact)
if subjects:
data['scraped_subjects'] = subjects
# add loaded votes to data
bill_votes = votes.pop((data['chamber'], data['session'],
data['bill_id']), [])
data['votes'].extend(bill_votes)
bill = db.bills.find_one({'state': data['state'],
'session': data['session'],
'chamber': data['chamber'],
'bill_id': data['bill_id']})
vote_matcher = VoteMatcher(data['state'])
if bill:
vote_matcher.learn_vote_ids(bill['votes'])
vote_matcher.set_vote_ids(data['votes'])
# match sponsor leg_ids
for sponsor in data['sponsors']:
id = get_legislator_id(state, data['session'], None,
sponsor['name'])
sponsor['leg_id'] = id
for vote in data['votes']:
# committee_ids
if 'committee' in vote:
committee_id = get_committee_id(state,
vote['chamber'],
vote['committee'])
vote['committee_id'] = committee_id
# vote leg_ids
for vtype in ('yes_votes', 'no_votes', 'other_votes'):
svlist = []
for svote in vote[vtype]:
id = get_legislator_id(state, data['session'],
vote['chamber'], svote)
svlist.append({'name': svote, 'leg_id': id})
vote[vtype] = svlist
data['_term'] = sessions[data['session']]
# Merge any version titles into the alternate_titles list
alt_titles = set(data.get('alternate_titles', []))
for version in data['versions']:
if 'title' in version:
alt_titles.add(version['title'])
if '+short_title' in version:
alt_titles.add(version['+short_title'])
try:
# Make sure the primary title isn't included in the
# alternate title list
alt_titles.remove(data['title'])
except KeyError:
pass
data['alternate_titles'] = list(alt_titles)
if not bill:
data['_keywords'] = list(bill_keywords(data))
insert_with_id(data)
else:
data['_keywords'] = list(bill_keywords(data))
update(bill, data, db.bills)
print 'imported %s bill files' % len(paths)
for remaining in votes.keys():
print 'Failed to match vote %s %s %s' % tuple([
r.encode('ascii', 'replace') for r in remaining])
populate_current_fields(state)
ensure_indexes()
def import_bills(state, data_dir):
data_dir = os.path.join(data_dir, state)
pattern = os.path.join(data_dir, 'bills', '*.json')
meta = db.metadata.find_one({'_id': state})
# Build a session to term mapping
sessions = {}
for term in meta['terms']:
for session in term['sessions']:
sessions[session] = term['name']
paths = glob.glob(pattern)
for path in paths:
with open(path) as f:
data = prepare_obj(json.load(f))
# clean up bill_id
data['bill_id'] = fix_bill_id(data['bill_id'])
subjects = data.pop('subjects', None)
if subjects:
data['scraped_subjects'] = subjects
bill = db.bills.find_one({'state': data['state'],
'session': data['session'],
'chamber': data['chamber'],
'bill_id': data['bill_id']})
for sponsor in data['sponsors']:
id = get_legislator_id(state, data['session'], None,
sponsor['name'])
sponsor['leg_id'] = id
for vote in data['votes']:
if 'committee' in vote:
committee_id = get_committee_id(state,
vote['chamber'],
vote['committee'])
vote['committee_id'] = committee_id
for vtype in ('yes_votes', 'no_votes', 'other_votes'):
svlist = []
for svote in vote[vtype]:
id = get_legislator_id(state, data['session'],
vote['chamber'], svote)
svlist.append({'name': svote, 'leg_id': id})
vote[vtype] = svlist
data['_term'] = sessions[data['session']]
# Merge any version titles into the alternate_titles list
alt_titles = set(data.get('alternate_titles', []))
for version in data['versions']:
if 'title' in version:
alt_titles.add(version['title'])
if '+short_title' in version:
alt_titles.add(version['+short_title'])
try:
# Make sure the primary title isn't included in the
# alternate title list
alt_titles.remove(data['title'])
except KeyError:
pass
data['alternate_titles'] = list(alt_titles)
if not bill:
data['_keywords'] = list(bill_keywords(data))
insert_with_id(data)
else:
data['_keywords'] = list(bill_keywords(data))
update(bill, data, db.bills)
print 'imported %s bill files' % len(paths)
populate_current_fields(state)
ensure_indexes()
def import_legislator(data):
data = prepare_obj(data)
if data.get('_scraped_name') is None:
data['_scraped_name'] = data['full_name']
# Rename 'role' -> 'type'
for role in data['roles']:
if 'role' in role:
role['type'] = role.pop('role')
# copy over LEVEL_FIELD into role
if settings.LEVEL_FIELD in data:
role[settings.LEVEL_FIELD] = data[settings.LEVEL_FIELD]
scraped_role = data['roles'][0]
scraped_term = scraped_role['term']
abbr = data[settings.LEVEL_FIELD]
spec = {
settings.LEVEL_FIELD: abbr,
'type': scraped_role['type'],
'term': scraped_term
}
if 'district' in scraped_role:
spec['district'] = scraped_role['district']
if 'chamber' in scraped_role:
spec['chamber'] = scraped_role['chamber']
# find matching legislator in current term
leg = db.legislators.find_one({
settings.LEVEL_FIELD: abbr,
'_scraped_name': data['_scraped_name'],
'roles': {
'$elemMatch': spec
}
})
# legislator with a matching old_role
if not leg:
spec.pop('term')
leg = db.legislators.find_one({
settings.LEVEL_FIELD: abbr,
'_scraped_name': data['_scraped_name'],
'old_roles.%s' % scraped_term: {
'$elemMatch': spec
}
})
if leg:
if 'old_roles' not in data:
data['old_roles'] = leg.get('old_roles', {})
# put scraped roles into their old_roles
data['old_roles'][scraped_term] = data['roles']
data['roles'] = leg['roles'] # don't overwrite their current roles
# active matching legislator from different term
if not leg:
spec.pop('term', None)
leg = db.legislators.find_one({
settings.LEVEL_FIELD: abbr,
'_scraped_name': data['_scraped_name'],
'roles': {
'$elemMatch': spec
}
})
if leg:
if 'old_roles' not in data:
data['old_roles'] = leg.get('old_roles', {})
# scraped_term < leg's term
if term_older_than(abbr, scraped_term, leg['roles'][0]['term']):
# move scraped roles into old_roles
data['old_roles'][scraped_term] = data['roles']
data['roles'] = leg['roles']
else:
data['old_roles'][leg['roles'][0]['term']] = leg['roles']
data = apply_filters(filters, data)
if leg:
update(leg, data, db.legislators)
return "update"
else:
insert_with_id(data)
return "insert"
def import_events(abbr, data_dir, import_actions=False):
data_dir = os.path.join(data_dir, abbr)
pattern = os.path.join(data_dir, 'events', '*.json')
for path in glob.iglob(pattern):
with open(path) as f:
data = prepare_obj(json.load(f))
def _resolve_ctty(committee):
return get_committee_id(data[settings.LEVEL_FIELD],
committee['chamber'],
committee['participant'])
def _resolve_leg(leg):
chamber = leg['chamber'] if leg['chamber'] in ['upper', 'lower'] \
else None
return get_legislator_id(abbr,
data['session'],
chamber,
leg['participant'])
resolvers = {
"committee": _resolve_ctty,
"legislator": _resolve_leg
}
for entity in data['participants']:
type = entity['participant_type']
id = None
if type in resolvers:
id = resolvers[type](entity)
else:
logger.warning("I don't know how to resolve a %s" % type)
entity['id'] = id
for bill in data['related_bills']:
bill_id = bill['bill_id']
bill_id = fix_bill_id(bill_id)
db_bill = db.bills.find_one({
"$or": [
{
settings.LEVEL_FIELD: abbr,
'session': data['session'],
'bill_id': bill_id
},
{
settings.LEVEL_FIELD: abbr,
'session': data['session'],
'alternate_bill_ids': bill_id
}
]
})
if not db_bill:
logger.warning("Error: Can't find %s" % bill_id)
db_bill = {}
db_bill['_id'] = None
# Events are really hard to pin to a chamber. Some of these are
# also a committee considering a bill from the other chamber, or
# something like that.
bill['id'] = db_bill['_id']
import_event(data)
ensure_indexes()
def import_committees(state, data_dir):
data_dir = os.path.join(data_dir, state)
pattern = os.path.join(data_dir, 'committees', '*.json')
meta = db.metadata.find_one({'_id': state})
current_term = meta['terms'][-1]['name']
current_session = meta['terms'][-1]['sessions'][-1]
paths = glob.glob(pattern)
for committee in db.committees.find({'state': state}):
committee['members'] = []
db.committees.save(committee)
if not paths:
# Not standalone committees
for legislator in db.legislators.find({
'roles': {'$elemMatch': {'term': current_term,
'state': state}}}):
for role in legislator['roles']:
if (role['type'] == 'committee member' and
'committee_id' not in role):
spec = {'state': role['state'],
'chamber': role['chamber'],
'committee': role['committee']}
if 'subcommittee' in role:
spec['subcommittee'] = role['subcommittee']
committee = db.committees.find_one(spec)
if not committee:
committee = spec
committee['_type'] = 'committee'
committee['members'] = []
committee['sources'] = []
if 'subcommittee' not in committee:
committee['subcommittee'] = None
insert_with_id(committee)
for member in committee['members']:
if member['leg_id'] == legislator['leg_id']:
break
else:
committee['members'].append(
{'name': legislator['full_name'],
'leg_id': legislator['leg_id'],
'role': role.get('position') or 'member'})
db.committees.save(committee, safe=True)
role['committee_id'] = committee['_id']
db.legislators.save(legislator, safe=True)
for path in paths:
with open(path) as f:
data = prepare_obj(json.load(f))
spec = {'state': state,
'chamber': data['chamber'],
'committee': data['committee']}
if 'subcommittee' in data:
spec['subcommittee'] = data['subcommittee']
committee = db.committees.find_one(spec)
if not committee:
insert_with_id(data)
committee = data
else:
update(committee, data, db.committees)
for member in committee['members']:
if not member['name']:
continue
leg_id = get_legislator_id(state, current_session,
data['chamber'],
member['name'])
if not leg_id:
print "No matches for %s" % member['name'].encode(
'ascii', 'ignore')
member['leg_id'] = None
continue
legislator = db.legislators.find_one({'_id': leg_id})
member['leg_id'] = leg_id
for role in legislator['roles']:
if (role['type'] == 'committee member' and
role['term'] == current_term and
role['committee_id'] == committee['_id']):
break
else:
new_role = {'type': 'committee member',
'committee': committee['committee'],
'term': current_term,
'chamber': committee['chamber'],
'committee_id': committee['_id'],
'state': state}
if 'subcommittee' in committee:
new_role['subcommittee'] = committee['subcommittee']
legislator['roles'].append(new_role)
legislator['updated_at'] = datetime.datetime.utcnow()
db.legislators.save(legislator, safe=True)
db.committees.save(committee, safe=True)
print 'imported %s committee files' % len(paths)
link_parents(state)
ensure_indexes()
def import_legislator(data):
data = prepare_obj(data)
if data.get('_scraped_name') is None:
data['_scraped_name'] = data['full_name']
# Rename 'role' -> 'type'
for role in data['roles']:
if 'role' in role:
role['type'] = role.pop('role')
# copy over LEVEL_FIELD into role
if settings.LEVEL_FIELD in data:
role[settings.LEVEL_FIELD] = data[settings.LEVEL_FIELD]
scraped_role = data['roles'][0]
scraped_term = scraped_role['term']
abbr = data[settings.LEVEL_FIELD]
spec = {settings.LEVEL_FIELD: abbr,
'type': scraped_role['type'],
'term': scraped_term}
if 'district' in scraped_role:
spec['district'] = scraped_role['district']
if 'chamber' in scraped_role:
spec['chamber'] = scraped_role['chamber']
# find matching legislator in current term
leg = db.legislators.find_one(
{settings.LEVEL_FIELD: abbr,
'_scraped_name': data['_scraped_name'],
'roles': {'$elemMatch': spec}})
# legislator with a matching old_role
if not leg:
spec.pop('term')
leg = db.legislators.find_one({
settings.LEVEL_FIELD: abbr,
'_scraped_name': data['_scraped_name'],
'old_roles.%s' % scraped_term: {'$elemMatch': spec}
})
if leg:
if 'old_roles' not in data:
data['old_roles'] = leg.get('old_roles', {})
# put scraped roles into their old_roles
data['old_roles'][scraped_term] = data['roles']
data['roles'] = leg['roles'] # don't overwrite their current roles
# active matching legislator from different term
if not leg:
spec.pop('term', None)
leg = db.legislators.find_one(
{settings.LEVEL_FIELD: abbr,
'_scraped_name': data['_scraped_name'],
'roles': {'$elemMatch': spec}})
if leg:
if 'old_roles' not in data:
data['old_roles'] = leg.get('old_roles', {})
# scraped_term < leg's term
if term_older_than(abbr, scraped_term, leg['roles'][0]['term']):
# move scraped roles into old_roles
data['old_roles'][scraped_term] = data['roles']
data['roles'] = leg['roles']
else:
data['old_roles'][leg['roles'][0]['term']] = leg['roles']
data = apply_filters(filters, data)
if leg:
update(leg, data, db.legislators)
return "update"
else:
insert_with_id(data)
return "insert"
