def toc_pagesets(self, data, facets):
# our primary facet is RPUBL.rattsfallspublikation, but we
# need to create one pageset for each value thereof.
pagesetdict = {}
selector_values = {}
facet = facets[0] # should be the RPUBL.rattsfallspublikation one
for row in data:
pagesetid = row['rpubl_rattsfallspublikation']
if pagesetid not in pagesetdict:
# Get the preferred court label from our own mapping,
# fall back to the skos:prefLabel of the publikation
label = self._rattsfallspublikation_label.get(
row['rpubl_rattsfallspublikation'],
Facet.resourcelabel(row, 'rpubl_rattsfallspublikation',
self.commondata))
pagesetdict[pagesetid] = TocPageset(label=label,
predicate=pagesetid,
pages=[])
selected = row['rpubl_arsutgava']
selector_values[(pagesetid, selected)] = True
for (pagesetid, value) in sorted(list(selector_values.keys()), reverse=True):
pageset = pagesetdict[pagesetid]
pageset.pages.append(TocPage(linktext=value,
title="Rättsfall från %s under %s" % (pageset.label, value),
binding=util.uri_leaf(pagesetid),
value=value))
# make sure pagesets are returned in the preferred, arbitrary order specified by _rattsfallspublikation_order
for x in pagesetdict.values():
assert x.label in self._rattsfallspublikation_order, "%s not in _rattsfallspublikation_order" % x.label
return sorted(list(pagesetdict.values()), key=lambda x: self._rattsfallspublikation_order.index(x.label))
def import_dataset(sourcegraph, targetgraph):
# print("Adding triples in %s to targetgraph" % filename)
# sourcegraph = rdflib.Graph()
# sourcegraph.parse(open(filename), format="n3")
# iterate through all named things (using skos:prefLabel)
for (sourceuri, name) in sourcegraph.subject_objects(predicate=SKOS.prefLabel):
targeturi = targetgraph.value(predicate=SKOS.prefLabel, object=name)
if not targeturi:
slug = sourcegraph.value(sourceuri, SKOS.altLabel)
if not slug:
print("WARNING: Can't find skos:altLabel for %s, using alternate method" % sourceuri)
slug = util.uri_leaf(str(sourceuri))
uri = "https://lagen.nu/dataset/%s" % str(slug).lower().translate(TRANS)
print(" Adding new resource %s" %uri)
targeturi = rdflib.URIRef(uri)
for (p, o) in sourcegraph.predicate_objects(subject=sourceuri):
if not targetgraph.value(targeturi, p): # we don't know the value for this pred
if p == DCTERMS.publisher:
o = URIMAP[o]
print(" Adding: %s %s %s" % (targeturi, sourcegraph.qname(p), o))
targetgraph.add((targeturi, p, o))
# finally add owl:sameAs if not already there
if sourceuri not in targetgraph.objects(targeturi, OWL.sameAs):
targetgraph.add((targeturi, OWL.sameAs, sourceuri))
print(" Asserting res %s owl:sameAs %s " % (targeturi, sourceuri))
URIMAP[sourceuri] = targeturi
def add_finegrained(desc, template, abbrslug):
# now create ~10 bunch of fine-grained templates for each
# fs-template that can mint uris for sections, paragraphs
# etc.
# "#K{kapnr}",
# "#K{kapnr}P{parnr}"
# "#K{kapnr}P{parnr}S{stnr}"
# "#K{kapnr}P{parnr}S{stnr}N{pnr}"
# "#P{parnr}"
# "#P{parnr}S{stnr}"
# "#P{parnr}S{stnr}N{pnr}"
# "#S{stnr}"
# "#S{stnr}N{pnr}"
proptuples = [(RPUBL.kapitelnummer, "K"),
(RPUBL.paragrafnummer, "P"),
(RINFOEX.styckenummer, "S"),
(RINFOEX.punktnummer, "N")]
while len(proptuples) > 1:
bindings = [RPUBL.forfattningssamling, RPUBL.arsutgava, RPUBL.lopnummer]
uritemplate = template + "#"
for p, fragletter in proptuples:
bindings.append(p)
with desc.rel(COIN.template):
uritemplate += fragletter + "{" + util.uri_leaf(p) + "}"
# print("adding uritemplate %s" % uritemplate)
desc.value(COIN.uriTemplate, uritemplate)
add_bindings(desc, bindings, abbrslug)
proptuples.pop(0)
def add_bindings(desc, bindings, slugFrom):
for b in bindings:
with desc.rel(COIN.binding):
desc.rel(COIN.property, b)
if b == RPUBL.forfattningssamling:
desc.value(COIN.variable, "fs")
desc.rel(COIN.slugFrom, slugFrom)
elif b == RDF.type:
desc.value(COIN.variable, "rtype")
desc.rel(COIN.slugFrom, slugFrom)
else:
desc.value(COIN.variable, util.uri_leaf(b))
def term(cls, row, binding='dcterms_publisher', resource_graph=None):
"""Returns the leaf part of the URI found in ``row[binding]``.
>>> row = {"rdf_type": "http://purl.org/ontology/bibo/Book",
... "dcterms_title": "A Tale of Two Cities",
... "dcterms_issued": "1859-04-30",
... "dcterms_publisher": "http://example.org/chapman_hall",
... "schema_free": "true"}
>>> Facet.term(row, "dcterms_publisher")
'chapman_hall'
"""
ret = util.uri_leaf(row[binding])
if not ret:
# FIXME: get a logger and complain. but also get something
# that can act as a URI fragmentx
ret = row[binding].replace(" ", "_")
return ret
def term(cls, row, binding='dcterms_publisher', resource_graph=None):
"""Returns the leaf part of the URI found in ``row[binding]``.
>>> row = {"rdf_type": "http://purl.org/ontology/bibo/Book",
... "dcterms_title": "A Tale of Two Cities",
... "dcterms_issued": "1859-04-30",
... "dcterms_publisher": "http://example.org/chapman_hall",
... "schema_free": "true"}
>>> Facet.term(row, "dcterms_publisher")
'chapman_hall'
"""
ret = util.uri_leaf(row[binding])
if not ret:
# FIXME: get a logger and complain. but also get something
# that can act as a URI fragmentx
ret = row[binding].replace(" ", "_")
return ret
def ident(row, binding, extra):
rdftype = row[binding]
if rdftype == str(self.ns['rpubl'].Utredningsbetankande):
if row['rpubl_utrSerie']:
leaf = util.uri_leaf(row['rpubl_utrSerie'])
if leaf.startswith("ds"):
return "ds"
elif leaf.startswith("sou"):
return "sou"
else:
assert leaf in ("sou", "ds"), "leaf was %s, unsure whether this is a SOU or a Ds." % leaf
else:
self.log.error("Row for %s is rpubl:Utredning but lacks rpubl:utrSerie" % row['uri'])
elif rdftype == str(self.ns['rpubl'].Kommittedirektiv):
return "dir"
elif rdftype == str(self.ns['rpubl'].Proposition):
return "prop"
else:
pass
def ident(row, binding, extra):
rdftype = row[binding]
if rdftype == str(self.ns['rpubl'].Utredningsbetankande):
if row['rpubl_utrSerie']:
leaf = util.uri_leaf(row['rpubl_utrSerie'])
if leaf.startswith("ds"):
return "ds"
elif leaf.startswith("sou"):
return "sou"
else:
assert leaf in (
"sou", "ds"
), "leaf was %s, unsure whether this is a SOU or a Ds." % leaf
else:
self.log.error(
"Row for %s is rpubl:Utredning but lacks rpubl:utrSerie"
% row['uri'])
elif rdftype == str(self.ns['rpubl'].Kommittedirektiv):
return "dir"
elif rdftype == str(self.ns['rpubl'].Proposition):
return "prop"
else:
pass
def toc_select_for_pages(self, data, pagesets, facets):
def idkey(row):
k = util.split_numalpha(row['dcterms_identifier'])
if " not " in row['dcterms_identifier']:
k[0] = "~" + k[0] # ensure notisfall sorts last
return k
facet = facets[0]
res = {}
documents = {}
for row in data:
key = facet.selector(row, None)
if key not in documents:
documents[key] = []
documents[key].append(row)
pagesetdict = {}
for pageset in pagesets:
pagesetdict[util.uri_leaf(pageset.predicate)] = pageset
for (binding, value) in sorted(documents.keys()):
pageset = pagesetdict[binding]
s = sorted(documents[(binding, value)], key=idkey)
res[(binding, value)] = [self.toc_item(binding, row)
for row in s]
return res
def mainfs(row, binding, resource_graph):
uri = URIRef(row[binding])
mainuri = resource_graph.value(uri, DCTERMS.isReplacedBy)
if mainuri:
uri = mainuri
return util.uri_leaf(uri)
def stats_slice(self, data, facet, resource_graph):
binding = resource_graph.qname(facet.rdftype).replace(":", "_")
if facet.dimension_label:
dimension_label = facet.dimension_label
elif self.config.legacyapi:
dimension_label = util.uri_leaf(str(facet.rdftype))
else:
dimension_label = binding
dimension_type = facet.dimension_type
if (self.config.legacyapi and dimension_type == "value"):
# legacyapi doesn't support the value type, we must
# convert it into ref, and convert all string values to
# fake resource ref URIs
dimension_type = "ref"
transformer = lambda x: ("http://example.org/fake-resource/%s" % x
).replace(" ", "_")
elif self.config.legacyapi and dimension_type == "term":
# legacyapi expects "Standard" over "bibo:Standard", which is what
# Facet.qname returns
transformer = lambda x: x.split(":")[1]
else:
transformer = lambda x: x
observations = Counter()
# one file per uri+observation seen -- avoid
# double-counting
observed = {}
for row in data:
observation = None
try:
# maybe if facet.dimension_type == "ref", selector
# should always be Facet.defaultselector? NOTE:
# we look at facet.dimension_type, not
# dimension_type, as the latter may be altered if
# legacyapi == True
if facet.dimension_type == "ref":
observation = transformer(
Facet.defaultselector(row, binding))
else:
observation = transformer(
facet.selector(row, binding, resource_graph))
except Exception as e:
# most of the time, we should swallow this
# exception since it's a selector that relies on
# information that is just not present in the rows
# from some repos. I think.
if hasattr(facet.selector, 'im_self'):
# try to find the location of the selector
# function for easier debugging
fname = "%s.%s.%s" % (facet.selector.__module__,
facet.selector.im_self.__name__,
facet.selector.__name__)
else:
# probably a lambda function
fname = facet.selector.__name__
# FIXME: do we need the repo name here to provide useful
# messages?
# self.log.warning("facet %s (%s) fails for row %s : %s %s" % (binding, fname, row['uri'], e.__class__.__name__, str(e)))
pass
if observation is not None:
k = (dimension_type, observation)
if (row['uri'], observation) not in observed:
observed[(row['uri'], observation)] = True
observations[k] += 1
return dimension_label, observations
def stats_slice(self, data, facet, resource_graph):
binding = resource_graph.qname(facet.rdftype).replace(":", "_")
if facet.dimension_label:
dimension_label = facet.dimension_label
elif self.config.legacyapi:
dimension_label = util.uri_leaf(str(facet.rdftype))
else:
dimension_label = binding
dimension_type = facet.dimension_type
if (self.config.legacyapi and
dimension_type == "value"):
# legacyapi doesn't support the value type, we must
# convert it into ref, and convert all string values to
# fake resource ref URIs
dimension_type = "ref"
transformer = lambda x: (
"http://example.org/fake-resource/%s" %
x).replace(
" ",
"_")
elif self.config.legacyapi and dimension_type == "term":
# legacyapi expects "Standard" over "bibo:Standard", which is what
# Facet.qname returns
transformer = lambda x: x.split(":")[1]
else:
transformer = lambda x: x
observations = Counter()
# one file per uri+observation seen -- avoid
# double-counting
observed = {}
for row in data:
observation = None
try:
# maybe if facet.dimension_type == "ref", selector
# should always be Facet.defaultselector? NOTE:
# we look at facet.dimension_type, not
# dimension_type, as the latter may be altered if
# legacyapi == True
if facet.dimension_type == "ref":
observation = transformer(Facet.defaultselector(
row, binding))
else:
observation = transformer(
facet.selector(
row,
binding,
resource_graph))
except Exception as e:
# most of the time, we should swallow this
# exception since it's a selector that relies on
# information that is just not present in the rows
# from some repos. I think.
if hasattr(facet.selector, 'im_self'):
# try to find the location of the selector
# function for easier debugging
fname = "%s.%s.%s" % (facet.selector.__module__,
facet.selector.im_self.__name__,
facet.selector.__name__)
else:
# probably a lambda function
fname = facet.selector.__name__
# FIXME: do we need the repo name here to provide useful
# messages?
# self.log.warning("facet %s (%s) fails for row %s : %s %s" % (binding, fname, row['uri'], e.__class__.__name__, str(e)))
pass
if observation is not None:
k = (dimension_type, observation)
if (row['uri'], observation) not in observed:
observed[(row['uri'], observation)] = True
observations[k] += 1
return dimension_label, observations
def mainfs(row, binding, resource_graph):
uri = URIRef(row[binding])
mainuri = resource_graph.value(uri, DCTERMS.isReplacedBy)
if mainuri:
uri = mainuri
return util.uri_leaf(uri)