def next_object(self, element):
"""Return the next DataOU object."""
result = DataOU()
# IVR 2007-12-24 FIXME: One of the attributes is special, and tags the
# OU's intended usage code (bruksområde). Find out which attribute
# this is.
# Iterate over *all* subelements
for sub in element.getiterator():
value = None
if sub.text:
value = sub.text.strip().encode("latin1")
if sub.tag == "Stedkode":
sko = make_sko(value)
if sko is not None:
result.add_id(self.tag2type[sub.tag], sko)
else:
# invalid value for the <Stedkode> tag
if self.logger:
self.logger.warn(
'Detected XML <Stedkode> '
'tag with invalid value: %s', value)
elif sub.tag == "Overordnetstedkode":
sko = make_sko(value)
if sko is not None:
result.parent = (result.NO_SKO, make_sko(value))
elif sub.tag == "Navn":
for name in self._make_names(sub):
result.add_name(name)
elif sub.tag in ("Adresse", ):
result.add_address(self._make_address(sub))
elif sub.tag in ("Startdato", "Sluttdato"):
date = self._make_mxdate(sub.text, format="%Y-%m-%d")
if sub.tag == "Startdato":
result.start_date = date
else:
result.end_date = date
# Whether the OU can be published in various online directories
result.publishable = False
for tmp in element.findall(".//Bruksomrade/Type"):
if tmp.text == "Tillatt Organisasjon":
result.publishable = True
# <StedType> tell us how an OU can be used. This information is
# represented in Cerebrum with the help of spreads and can be
# accessed via the EntitySpread interface.
result.add_usage_code(tmp.text)
celems = element.findall("Kommunikasjon")
for sub in celems:
ct = self._make_contact(sub)
if ct:
result.add_contact(ct)
# We require an OU to have a name.
# Ideally, the information about expired OUs should be complete as
# well, but we won't be this lucky in our lifetimes. So, for expired
# OUs we won't care about the names.
# Neither do we care about the missing names of not yet active
# OUs; we choose to hope that the names will be in place when
# the OU becomes active.
if result.get_name(DataOU.NAME_LONG) is None:
ou_no_sko_str = result.get_id(DataOU.NO_SKO)
if not ou_no_sko_str:
ou_no_sko_str = 'Missing a valid NO_SKO value'
if result.end_date and result.end_date < now():
if self.logger:
self.logger.debug("No name for expired OU %s",
ou_no_sko_str)
elif result.start_date and result.start_date > now():
if self.logger:
self.logger.debug("No name for future OU %s",
ou_no_sko_str)
else:
if self.logger:
self.logger.warn("No name available for OU %s",
ou_no_sko_str)
return None
return result
def next_object(self, element):
def get_value(element_value):
return ensure_unicode(element_value, self.encoding)
def extract(element_attr):
return get_value(element.get(element_attr, ""))
result = DataOU()
# A lot of data is buried in attributes
# Own ID -- sko
sko = tuple([int(element.get(x)) for x in ("fakultetnr",
"instituttnr",
"gruppenr")])
result.add_id(result.NO_SKO, sko)
# Parent ID -- sko
sko = tuple([int(element.get(x)) for x in ("fakultetnr_for_org_sted",
"instituttnr_for_org_sted",
"gruppenr_for_org_sted")])
result.parent = (result.NO_SKO, sko)
# Some weird ID
if element.get("nsd_kode"):
result.add_id(result.NO_NSD,
get_value(element.get("nsd_kode")))
# Activity period
result.start_date = self._make_mxdate(element.get("dato_opprettet"))
result.end_date = self._make_mxdate(element.get("dato_nedlagt"))
# Accessibility for catalogues
if element.get("opprettetmerke_for_oppf_i_kat"):
result.publishable = True
for name_kind, candidates, lang in ((result.NAME_LONG,
("stedlangnavn_bokmal",
"stedkortnavn_bokmal",
"stednavnfullt", "stednavn"),
"nb"),
(result.NAME_LONG,
("stedlangnavn_engelsk",
"stedkortnavn_engelsk"),
"en"),
(result.NAME_ACRONYM,
("akronym",),
"nb"),
(result.NAME_SHORT,
("forkstednavn",),
"nb")):
value = self._pull_name(element, *candidates)
if value:
result.add_name(DataName(name_kind, value, lang))
for (xmlkind, kind) in (("besok", DataAddress.ADDRESS_BESOK),
("intern", DataAddress.ADDRESS_POST)):
zip = extract("poststednr_%s_adr" % xmlkind)
street = None
if xmlkind == "intern":
try:
p_o_box = int(extract("stedpostboks"))
if p_o_box and int(zip) // 100 == 3:
street = "Postboks %d Blindern" % p_o_box
except ValueError:
pass
if street is None:
street = (extract("adresselinje1_%s_adr" % xmlkind),
extract("adresselinje2_%s_adr" % xmlkind))
result.add_address(
DataAddress(kind=kind,
street=street,
zip=zip,
city=extract("poststednavn_%s_adr" % xmlkind),
country=extract("landnavn_%s_adr" % xmlkind)))
# FIXME: priority assignment is a bit random at the moment.
priority = 0
for sub in element.findall("komm"):
ct = self._make_contact(sub)
if ct:
kind, value = ct
result.add_contact(DataContact(kind,
get_value(value),
priority))
priority += 1
return result
def next_object(self, element):
"""Return the next DataOU object."""
result = DataOU()
# IVR 2007-12-24 FIXME: One of the attributes is special, and tags the
# OU's intended usage code (bruksområde). Find out which attribute
# this is.
# Iterate over *all* subelements
for sub in element.getiterator():
value = None
if sub.text:
value = ensure_unicode(sub.text.strip(), self.encoding)
if sub.tag == "Stedkode":
sko = make_sko(value)
if sko is not None:
result.add_id(self.tag2type[sub.tag], sko)
else:
# invalid value for the <Stedkode> tag
if self.logger:
self.logger.warn(
'Detected XML <Stedkode> '
'tag with invalid value: %s',
value
)
elif sub.tag == "Overordnetstedkode":
sko = make_sko(value)
if sko is not None:
result.parent = (result.NO_SKO, sko)
elif sub.tag == "Navn":
for name in self._make_names(sub):
result.add_name(name)
elif sub.tag in ("Adresse",):
result.add_address(self._make_address(sub))
elif sub.tag in ("Startdato", "Sluttdato"):
date = self._make_mxdate(sub.text, format="%Y-%m-%d")
if sub.tag == "Startdato":
result.start_date = date
else:
result.end_date = date
# Whether the OU can be published in various online directories
result.publishable = False
for tmp in element.findall(".//Bruksomrade/Type"):
if tmp.text == "Tillatt Organisasjon":
result.publishable = True
# <StedType> tell us how an OU can be used. This information is
# represented in Cerebrum with the help of spreads and can be
# accessed via the EntitySpread interface.
result.add_usage_code(tmp.text)
celems = element.findall("Kommunikasjon")
for sub in celems:
ct = self._make_contact(sub)
if ct:
result.add_contact(ct)
# We require an OU to have a name.
# Ideally, the information about expired OUs should be complete as
# well, but we won't be this lucky in our lifetimes. So, for expired
# OUs we won't care about the names.
# Neither do we care about the missing names of not yet active
# OUs; we choose to hope that the names will be in place when
# the OU becomes active.
if result.get_name(DataOU.NAME_LONG) is None:
ou_no_sko_str = result.get_id(DataOU.NO_SKO)
if not ou_no_sko_str:
ou_no_sko_str = 'Missing a valid NO_SKO value'
if result.end_date and result.end_date < now():
if self.logger:
self.logger.debug("No name for expired OU %s",
ou_no_sko_str)
elif result.start_date and result.start_date > now():
if self.logger:
self.logger.debug("No name for future OU %s",
ou_no_sko_str)
else:
if self.logger:
self.logger.warn("No name available for OU %s",
ou_no_sko_str)
return None
return result
def next_object(self, element):
def get_value(element_value):
return ensure_unicode(element_value, self.encoding)
def extract(element_attr):
return get_value(element.get(element_attr, ""))
result = DataOU()
# A lot of data is buried in attributes
# Own ID -- sko
sko = tuple([
int(element.get(x))
for x in ("fakultetnr", "instituttnr", "gruppenr")
])
result.add_id(result.NO_SKO, sko)
# Parent ID -- sko
sko = tuple([
int(element.get(x))
for x in ("fakultetnr_for_org_sted", "instituttnr_for_org_sted",
"gruppenr_for_org_sted")
])
result.parent = (result.NO_SKO, sko)
# Some weird ID
if element.get("nsd_kode"):
result.add_id(result.NO_NSD, get_value(element.get("nsd_kode")))
# Activity period
result.start_date = self._make_mxdate(element.get("dato_opprettet"))
result.end_date = self._make_mxdate(element.get("dato_nedlagt"))
# Accessibility for catalogues
if element.get("opprettetmerke_for_oppf_i_kat"):
result.publishable = True
for name_kind, candidates, lang in ((result.NAME_LONG,
("stedlangnavn_bokmal",
"stedkortnavn_bokmal",
"stednavnfullt", "stednavn"),
"nb"), (result.NAME_LONG,
("stedlangnavn_engelsk",
"stedkortnavn_engelsk"),
"en"),
(result.NAME_ACRONYM,
("akronym", ), "nb"),
(result.NAME_SHORT,
("forkstednavn", ), "nb")):
value = self._pull_name(element, *candidates)
if value:
result.add_name(DataName(name_kind, value, lang))
for (xmlkind, kind) in (("besok", DataAddress.ADDRESS_BESOK),
("intern", DataAddress.ADDRESS_POST)):
zip = extract("poststednr_%s_adr" % xmlkind)
street = None
if xmlkind == "intern":
try:
p_o_box = int(extract("stedpostboks"))
if p_o_box and int(zip) // 100 == 3:
street = "Postboks %d Blindern" % p_o_box
except ValueError:
pass
if street is None:
street = (extract("adresselinje1_%s_adr" % xmlkind),
extract("adresselinje2_%s_adr" % xmlkind))
result.add_address(
DataAddress(kind=kind,
street=street,
zip=zip,
city=extract("poststednavn_%s_adr" % xmlkind),
country=extract("landnavn_%s_adr" % xmlkind)))
# FIXME: priority assignment is a bit random at the moment.
priority = 0
for sub in element.findall("komm"):
ct = self._make_contact(sub)
if ct:
kind, value = ct
result.add_contact(
DataContact(kind, get_value(value), priority))
priority += 1
return result