def test_normalize_body_name(self):
utils = Utils()
self.assertEqual("Bedburg", utils.normalize_body_name("Stadt Bedburg"))
self.assertEqual("Leipzig", utils.normalize_body_name("Leipzig"))
self.assertEqual(
"Bad Münstereifel", utils.normalize_body_name("Stadt Bad Münstereifel ")
)
class Cli:
"""Tries to import all required data (oparl system id, ags and city name for osm) from a user input
that might be the name of a city, the url of a body or the url of a system.
It uses the information from wikidata, open street map and the oparl mirror.
"""
def __init__(self):
self.index = [] # type: List[Tuple[str, str, str]]
self.utils = Utils()
def load_index(self) -> List[Tuple[str, str, str]]:
""" " Loads the list of known endpoints from the oparl mirror if it has not been loaded yet"""
if self.index:
return self.index
next_page = settings.OPARL_INDEX
while next_page:
response = requests_get(next_page)
data = response.json()
next_page = data["links"].get("next")
for body in data["data"]:
if not "oparl-mirror:originalId" in body:
continue
self.index.append((
body.get("name") or body["oparl-mirror:originalId"],
body["oparl-mirror:originalId"],
body["id"],
))
return self.index
def from_userinput(
self,
userinput: str,
mirror: bool,
ags: Optional[str],
skip_body_extra: bool = False,
skip_files: bool = False,
) -> None:
body_id, entrypoint = self.get_entrypoint_and_body(userinput, mirror)
importer = Importer(get_loader_from_system(entrypoint))
body_data, dotenv = self.import_body_and_metadata(
body_id, importer, userinput, ags, skip_body_extra)
logger.info("Loading the bulk data from the oparl api")
importer.fetch_lists_initial([body_data])
# Also avoid "MySQL server has gone away" errors due to timeouts
# https://stackoverflow.com/a/32720475/3549270
db.close_old_connections()
logger.info("Loading the data into the database")
importer.import_objects()
if not skip_files:
logger.info("Loading the files")
importer.load_files(fallback_city=userinput)
if dotenv:
logger.info(
f"Done! Please add the following line to your dotenv file: \n\n{dotenv}\n"
)
def import_body_and_metadata(
self,
body_id: str,
importer: Importer,
userinput: str,
ags: Optional[str],
skip_body_extra: bool = False,
) -> Tuple[JSON, str]:
logger.info(f"Fetching the body {body_id}")
[body_data] = importer.load_bodies(body_id)
logger.info("Importing the body")
[body] = importer.import_bodies()
importer.converter.default_body = body
logger.info("Looking up the Amtliche Gemeindeschlüssel")
if ags:
if len(ags) != 5 and len(ags) != 8:
logger.warning(
"Your Amtlicher Gemeindeschlüssel has {} digits instead of 5 or 8"
.format(len(ags)))
body.ags = ags
else:
ags, match_name = self.get_ags(body, importer.loader.system,
userinput)
body.ags = ags
# Sometimes there's a bad short name (e.g. "Rat" for Erkelenz),
# so we use the name that's in wikidata instead
body.short_name = match_name
body.save()
logger.info("Using {} as Amtliche Gemeindeschlüssel for '{}'".format(
body.ags, body.short_name))
dotenv = ""
if body.id != settings.SITE_DEFAULT_BODY:
dotenv += f"SITE_DEFAULT_BODY={body.id}\n"
if dotenv:
logger.info(
"Found the oparl endpoint. Please add the following line to your dotenv file "
"(you'll be reminded again after the import finished): \n\n" +
dotenv)
if not skip_body_extra:
logger.info("Importing the shape of the city")
import_outline(body)
logger.info("Importing the streets")
import_streets(body)
logger.info(
f"Body {body.short_name} import with geo data successful.")
else:
logger.info(
f"Body {body.short_name} import successful. "
f"Don't forget to run import_streets, import_amenities and import_outline"
)
return body_data.data, dotenv
def get_entrypoint_and_body(self,
userinput: str,
mirror: bool = False) -> Tuple[str, str]:
try:
URLValidator()(userinput)
is_url = True
except ValidationError:
is_url = False
if not is_url:
entrypoint, body_id = self.get_endpoint_from_cityname(
userinput, mirror)
else:
entrypoint, body_id = self.get_endpoint_from_body_url(userinput)
logger.info(
f"Your body id is {body_id} and your system id is {entrypoint}")
return body_id, entrypoint
def get_endpoint_from_body_url(self, userinput: str) -> Tuple[str, str]:
# We can't use the resolver here as we don't know the system url yet, which the resolver needs for determining
# the cache folder
logging.info(f"Using {userinput} as url")
response = requests_get(userinput)
data = response.json()
if data.get("type") not in [
"https://schema.oparl.org/1.0/Body",
"https://schema.oparl.org/1.1/Body",
]:
raise RuntimeError(
"The url you provided didn't point to an oparl body")
endpoint_system = data["system"]
endpoint_id = data["id"]
if userinput != endpoint_id:
logger.warning(
f"The body's url '{userinput}' doesn't match the body's id '{endpoint_id}'"
)
return endpoint_system, endpoint_id
def get_ags(self, body: Body, system: JSON,
userinput: str) -> Tuple[str, str]:
"""
This function tries:
1. The ags field in the oparl body
2. Querying wikidata with
a) the body's short name
b) the user's input
c) the body's full name
d) the system's name
e) locality in the location
Returns the ags and the name that did match
"""
ags = body.ags
if ags:
if len(ags) == 8 or len(ags) == 5:
return ags, body.short_name
else:
logger.error(
f"Ignoring ags '{ags}' with invalid length {len(ags)}")
district = bool(
re.match(settings.DISTRICT_REGEX, body.name, re.IGNORECASE))
to_check = [
("body short name", body.short_name),
("user input", userinput),
("body name", body.name),
]
if system.get("name"):
short_system_name = self.utils.normalize_body_name(system["name"])
to_check.append(("system name", short_system_name))
if body.center and body.center.locality:
locality = body.center.locality
to_check.append(("body location locality", locality))
for source, value in to_check:
ags = city_to_ags(value, district)
if ags:
logger.debug(f"Found ags using the {source}: '{value}'")
return ags, value
raise RuntimeError(
f"Could not determine the Amtliche Gemeindeschlüssel using {to_check}.\n"
f"Set it manually using `--ags`")
def get_endpoint_from_cityname(self, userinput: str,
mirror: bool) -> Tuple[str, str]:
matching = [] # type: List[Tuple[str, str, str]]
for (name, original_id, mirror_id) in self.load_index():
if userinput.casefold() in name.casefold():
# The oparl mirror doesn't give us the system id we need
response = requests_get(original_id)
system_id = response.json()["system"]
if mirror:
matching.append((name, system_id, mirror_id))
else:
matching.append((name, system_id, original_id))
if len(matching) == 0:
raise RuntimeError(
f"Could not find anything for '{userinput}'. "
f"Please check that it is in the list under {settings.OPARL_INDEX} or provide the body id."
)
if len(matching) > 1:
exact_matches = [
i for i in matching if i[0].casefold() == userinput.casefold()
]
if len(exact_matches) == 1:
matching = exact_matches
else:
logger.warning(
f"Found those entries: {json.dumps(matching, indent=4)}")
raise RuntimeError(
f"There are {len(matching)} matches and {len(exact_matches)} exact matches for '{userinput}' and "
"I can't decide which one to use. Please provide a body url yourself."
)
return matching[0][1:3]