def getAddressJson(eid):
# json with all geocoded data
j = {}
with db_old.getCursor() as cur:
cur = db_old.execute(cur, "SELECT json FROM entities WHERE eid=%s", [eid])
row = cur.fetchone()
if row is None: return None
j = json.loads(row["json"])
# TODO: do not duplicate this with code in verejne/
def getComponent(json, typeName):
try:
for component in json[0]["address_components"]:
if typeName in component["types"]:
return component["long_name"]
return ""
except:
return ""
# types description: https://developers.google.com/maps/documentation/geocoding/intro#Types
# street / city can be defined in multiple ways
address = {
"street": (
getComponent(j, "street_address") +
getComponent(j, "route") +
getComponent(j, "intersection") +
" " + getComponent(j, "street_number")
),
"city": getComponent(j, "locality"),
"zip": getComponent(j, "postal_code"),
"country": getComponent(j, "country"),
}
return address
def get(self):
# Parse input
try:
firstname = self.request.GET["firstname"].encode("utf8")
surname = self.request.GET["surname"].encode("utf8")
address = self.request.GET["address"].encode("utf8")
except:
self.abort(400, detail="Unable to parse input text")
# Carry-out the logic
q = """
SELECT DISTINCT
eid AS eid
FROM
entities
WHERE
to_tsvector('unaccent', entity_name) @@ plainto_tsquery('unaccent', %s)
AND
to_tsvector('unaccent', address) @@ plainto_tsquery('unaccent', %s)
LIMIT 20;
"""
with db_old.getCursor() as cur:
cur = db_old.execute(cur, q, (firstname + ' ' + surname, address))
result = []
for row in cur:
try:
result.append({"eid": row["eid"]})
except:
pass
self.returnJSON(result)
def get(self):
try:
eid = int(self.request.GET["eid"])
except:
self.abort(400,
detail="Could not parse parameter 'eid' as an integer")
# Copy total sum of contracts if present
result = {}
entities = webapp2.get_app().registry['entities']
if eid in entities.eid_to_index:
entity = entities.entities[entities.eid_to_index[eid]]
result["total_contracts"] = entity.contracts
# load info data from individual tables
data_sources = webapp2.get_app().registry['data_sources']
for table in data_sources:
print "Processing table", table
columns = ",".join(data_sources[table])
sql = "SELECT DISTINCT " + columns + " FROM entities"
if (table != "entities"):
sql += " JOIN " + table + " ON " + "entities.id=" + table + ".id"
sql += " WHERE entities.eid=%s"
with db_old.getCursor() as cur:
cur = db_old.execute(cur, sql, [eid])
current = []
for row in cur:
# TODO: is this needed?
r = {key: row[key] for key in row}
current.append(r)
result[table] = current
result["related"] = entities.getRelated(eid)
self.returnJSON(result)
def getForTable(table):
sql = "SELECT eid FROM " + table + \
" JOIN entities ON entities.id = " + table + ".id" + \
" WHERE ico = %s" + \
" LIMIT 1"
with db_old.getCursor() as cur:
cur = db_old.execute(cur, sql, [ico])
row = cur.fetchone()
if row is None: return None
return row["eid"]
def IcoToLatLngMap():
output_map = {}
for table in ["orsresd_data", "firmy_data", "new_orsr_data"]:
with db_old.getCursor() as cur:
sql = "SELECT ico, lat, lng FROM " + table + \
" JOIN entities on entities.id = " + table + ".id" + \
" WHERE ico IS NOT NULL"
db_old.execute(cur, sql)
for row in cur: output_map[int(row["ico"])] = (row["lat"], row["lng"])
return output_map
def loadExtra(self, sql, column, attribute, mapping=None):
log("loadExtra " + column + ", " + attribute)
with db.getCursor() as cur:
cur = db.execute(cur, sql)
for row in cur:
eid = row["eid"]
if not eid in self.eid_to_index: continue
value = row[column]
if mapping is not None: value = mapping(value)
setattr(self.entities[self.eid_to_index[eid]], attribute,
value)
def getRelated(self, eid):
with db.getCursor() as cur:
indices = set()
cur = db.execute(cur, "(select distinct eid1 as eid from related where eid2=%s and " \
+ "eid2!=eid1) union (select distinct eid2 as eid from related where eid1=%s and eid2!=eid1)", [eid, eid])
for row in cur:
indices.add(row["eid"])
result = [
RawJson(self.entities[self.eid_to_index[index]].json)
for index in indices if index in self.eid_to_index
]
return result
def getColumnForTableIco(table, column, ico):
sql = "SELECT " + column + " FROM " + table + \
" JOIN entities ON entities.id = " + table + ".id" + \
" WHERE ico = %s" + \
" LIMIT 1"
with db_old.getCursor() as cur:
try:
cur = db_old.execute(cur, sql, [ico])
row = cur.fetchone()
if row is None: return None
return row[column]
except:
return None
def get(self):
try:
text = self.request.GET["text"].encode("utf8")
except:
self.abort(400, detail="Unable to parse input text")
with db_old.getCursor() as cur:
sql = "SELECT DISTINCT eid AS eid FROM entities " + \
"WHERE to_tsvector('unaccent', entity_name) @@ plainto_tsquery('unaccent', %s) " + \
"LIMIT 20"
cur = db_old.execute(cur, sql, [text])
result = []
for row in cur:
try:
result.append({"eid": row["eid"]})
except:
pass
self.returnJSON(result)
def loadEntities(self, batch_size):
cur = db.getCursor()
db.execute(cur, "SELECT eid, entity_name, lat, lng, json, address " +
"FROM entities " +
"WHERE (entity_name IS NOT NULL) AND " +
" (entity_name != '') AND " +
" (has_data IS NOT NULL)" + \
"LIMIT %s", [batch_size])
for row in cur:
# if multiple rows share common eid, use only the first one
# TODO: use the most recent address
if (row["eid"] in self.eid_to_index): continue
j = json.loads(row["json"])
entity = Entity()
country = getComponent(j, "country")
entity.country = country
entity.slovakia = (entity.country == "Slovakia")
entity.eid = row["eid"]
entity.lat = float(row["lat"])
entity.lng = float(row["lng"])
entity.name = row["entity_name"].decode("utf8")
address = row["address"].decode("utf8")
if country is not None:
address = address.replace(", " + country, "")
psc = getComponent(j, "postal_code")
if psc is not None:
address = address.replace(psc + " ", "")
entity.address = address
entity.origin = 0
entity.size = 1
try:
city = getComponent(j, "locality")
entity.city = city
except:
entity.city = None
self.entities.append(entity)
num_entities = len(self.entities)
# Debug output entity json
if (num_entities % 30000 == 0):
print json.dumps(json.loads(row["json"]), indent=4)
self.eid_to_index[entity.eid] = num_entities
cur.close()