def test_mosdef_only(testresourcepath, expected_modout1):
modin = newmodel()
literate.parse(INPUT_GRAPH_1, modin)
modin = newmodel()
literate.parse(INPUT_GRAPH_1, modin)
FINGERPRINT_RULES = {
SCH_NS('MusicAlbum'):
(if_(contains(follow(SCH_NS('byArtist')), DOC_NS('md')),
materialize(COPY()))),
SCH_NS('Person'): (materialize(COPY())),
}
ppl = generic_pipeline(FINGERPRINT_RULES, {}, {})
modout = ppl.run(input_model=modin)
# Use -s to see this
print('=' * 10, 'test_mosdef_only', '=' * 10)
literate.write(modout)
# import pprint; pprint.pprint(list(iter(modout)))
assert len(modout) == 17
assert len(
list(util.all_origins(modout, only_types={SCH_NS('MusicAlbum')}))) == 1
assert len(list(util.all_origins(modout,
only_types={SCH_NS('Person')}))) == 3
def run(self, input_model=None, raw_source=None, output_model=None, **kwargs):
'''
Process an input, either an input Versa model or in some raw record format
through a sequence of transform stages, to generate a versa model of output resources
Caller must provide either an input_model or a raw_source, but can provide
any combination of these, depending on the expectations of the defined stages
Args:
input_model: Versa model which serves as the starting point
raw_source: raw input data, a possible optimization if it's impractical to directly
represent as a Versa model, but can be interpreted by the stages as if it were
output_model: optional output model, which might be provided to add transform results
to existing data, or to use a specialized Versa model implementation
kwargs: any additional parameters which are passed as they are to all the stages
Returns:
output_model: Same reference as the input output_model, if provided, otherwise a new
model containing the results of the transform
'''
self.check_update_stages()
self.input_model = newmodel() if input_model is None else input_model
self.output_model = newmodel() if output_model is None else output_model
self._raw_source = raw_source
self.fingerprints = {}
# First tuple item is just sortkey, so discarded
for _, stage in self._stages:
retval = stage(**kwargs)
if retval is False:
#Signal to abort
break
return self.output_model
def test_basics_1(testresourcepath, expected_modout1):
modin = newmodel()
modin_fpath = 'schemaorg/catcherintherye-ugly.md'
literate.parse(
open(os.path.join(testresourcepath, modin_fpath)).read(), modin)
FINGERPRINT_RULES = {
SCH_NS('Book'): (materialize(
BF_NS('Instance'),
fprint=[
(BF_NS('isbn'), follow(SCH_NS('isbn'))),
],
))
}
TRANSFORM_RULES = {
SCH_NS('name'):
link(rel=BF_NS('name')),
SCH_NS('author'):
materialize(BF_NS('Person'),
BF_NS('creator'),
vars={
'birthDate':
follow(SCH_NS('authorBirthDate'),
origin=var('input-resource'))
},
fprint=[
(BF_NS('name'), target()),
(BF_NS('birthDate'), var('birthDate')),
],
links=[
(BF_NS('name'), target()),
(BF_NS('birthDate'), var('birthDate')),
]),
}
modout = newmodel()
def new_entity_hook(eid):
# Add a triple to each materialized resource
modout.add(eid, 'http://example.org/materializedBy', 'py.test')
return
ctxextras = {'@new-entity-hook': new_entity_hook}
root_ctx = DUMMY_CONTEXT.copy(output_model=modout, extras=ctxextras)
ppl = generic_pipeline(FINGERPRINT_RULES,
TRANSFORM_RULES,
LABELIZE_RULES,
root_ctx=root_ctx)
ppl.run(input_model=modin, output_model=modout)
# Use -s to see this
print('=' * 10, 'test_basics_1', '=' * 10)
literate.write(modout)
assert len(
list(modout.match(None, 'http://example.org/materializedBy',
None))) == 2
def test_simpleobj_usecase1():
m = newmodel()
tmpl = Template(
'# http://example.org#{{ Wikidata }}\n\n * <http://example.org/voc/copyright>: {{ _["©"] }}'
)
# use -s option to see the nosy print
m = newmodel()
parse(objmock(), tmpl, m, nosy=print)
assert len(m) == 1, repr(m)
assert ('http://example.org#Q15761337', 'http://example.org/voc/copyright',
'2016', {}) == next(m.match())
def test_versa_syntax1(testresourcepath):
config = {
'autotype-h1': 'http://example.org/r1',
'autotype-h2': 'http://example.org/r2',
'interpretations': {
VERSA_BASEIRI + 'refines': VERSA_BASEIRI + 'resourceset',
VERSA_BASEIRI + 'properties': VERSA_BASEIRI + 'resourceset',
VERSA_BASEIRI + 'synonyms': VERSA_BASEIRI + 'resourceset'
}
}
m1 = newmodel(baseiri='http://example.org/')
# from_markdown(VERSA_LITERATE1, m, encoding='utf-8')
doc = open(os.path.join(testresourcepath, 'doc1.md')).read()
literate.parse(doc, m1, config=config)
m2 = newmodel(baseiri='http://example.org/')
# from_markdown(VERSA_LITERATE1, m, encoding='utf-8')
doc = open(os.path.join(testresourcepath, 'doc1.abbr.md')).read()
literate.parse(doc, m2, config=config)
# logging.debug('VERSA LITERATE EXAMPLE 1')
equiv_results = [list(m1.match()), list(m2.match())]
for results in equiv_results:
# import pprint; pprint.pprint(results)
assert len(results) == 6
assert (I('http://uche.ogbuji.net/ndewo/'),
I('http://bibfra.me/purl/versa/type'), 'http://example.org/r1',
{}) in results
assert (I('http://uche.ogbuji.net/ndewo/'),
I('http://www.w3.org/TR/html5/title'), 'Ndewo, Colorado', {
'@lang': None
}) in results
assert (I('http://uche.ogbuji.net/ndewo/'),
I('http://www.w3.org/TR/html5/link-type/author'),
I('http://uche.ogbuji.net/'), {
I('http://www.w3.org/TR/html5/link/description'):
'Uche Ogbuji'
}) in results
assert (
I('http://uche.ogbuji.net/ndewo/'),
I('http://www.w3.org/TR/html5/link-type/see-also'),
I('http://www.goodreads.com/book/show/18714145-ndewo-colorado'), {
I('http://www.w3.org/TR/html5/link/label'): 'Goodreads'
}) in results
assert (I('http://uche.ogbuji.net/'),
I('http://bibfra.me/purl/versa/type'), 'http://example.org/r1',
{}) in results
assert (I('http://uche.ogbuji.net/'),
I('http://www.w3.org/TR/html5/link-type/see-also'),
I('http://uche.ogbuji.net/ndewo/'), {}) in results
def main(source):
'Transform CSV SOURCE file to BF Lite in Versa'
ppl = csv_bibframe_pipeline()
input_model = newmodel()
with open(source) as csvfp:
for row_model in csv.parse_iter(csvfp, VLITERATE_TEMPLATE):
if row_model: input_model.update(row_model)
# Debug print of input model
# literate.write([input_model], out=sys.stdout)
output_model = ppl.run(input_model=input_model)
print('Low level JSON dump of output data model: ')
util.jsondump(output_model, sys.stdout)
print('\n') # 2 CRs
print('Versa literate form of output: ')
literate.write(output_model, out=sys.stdout)
print('Diagram from extracted a sample: ')
out_resources = []
for vs in ppl.fingerprints.values():
out_resources.extend(vs)
ITYPE = BF_NS('Instance')
instances = [
r for r in out_resources
if ITYPE in util.resourcetypes(output_model, r)
]
zoomed, _ = util.zoom_in(output_model, random.choice(instances), depth=2)
mermaid.write(zoomed)
def test_book_cases(label, transforms, asserter):
idg = idgen(EXAMPLE_ORG)
existing_ids = set()
out_m = newmodel()
rid = SIMPLE_BOOK['id']
out_m.add(rid, VTYPE_REL, BOOK_TYPE)
in_m = newmodel()
for k, v in SIMPLE_BOOK.items():
ctxlink = (rid, k, v, {})
func = transforms.get(k)
if func:
ctx = context(ctxlink, in_m, out_m, base=SCHEMA_ORG, idgen=idg)
func(ctx)
asserter(out_m)
def test_csv_usecase1():
m = newmodel()
tmpl = '# http://example.org#{Wikidata}\n\n * <http://example.org/voc/copyright>: {%C2%A9}'
# use -s option to see the nosy print
m = next(parse_iter(object(), tmpl, csv_fact=csvmock, nosy=print))
assert len(m) == 1, repr(m)
assert ('http://example.org#Q15761337', 'http://example.org/voc/copyright', '2016', {}) == next(m.match())
def Xtest_versa_syntax1():
# logging.debug(recs)
m = newmodel()
m.create_space()
# from_markdown(VERSA_LITERATE1, m, encoding='utf-8')
literate.parse(VERSA_LITERATE1, m)
logging.debug('VERSA LITERATE EXAMPLE 1')
for link in m.match():
logging.debug('Result: {0}'.format(repr(link)))
def test_parse1(ntrips_1):
m = newmodel()
parse(ntrips_1, m)
assert len(m) == 3, repr(m)
assert 'Dave Beckett' in [
t for (o, r, t, a) in m.match(NT_SPEC, DC_CREATOR)
]
assert 'Art Barstow' in [
t for (o, r, t, a) in m.match(NT_SPEC, DC_CREATOR)
]
assert W3C in [t for (o, r, t, a) in m.match(NT_SPEC, DC_PUBLISHER)]
def main(source):
'Transform CSV SOURCE file to Schema.org in Versa'
ppl = csv_schema_pipeline()
input_model = newmodel()
with open(source) as csvfp:
csv.parse(csvfp, VLITERATE_TEMPLATE, input_model)
# Debug print of input model
# literate.write([input_model], out=sys.stdout)
output_model = ppl.run(input_model=input_model)
print('Resulting record Fingerprints:', ppl.fingerprints)
print('Low level JSON dump of output data model: ')
util.jsondump(output_model, sys.stdout)
print('Versa literate form of output: ')
literate.write(output_model, out=sys.stdout)
def test_basics_1(testresourcepath, expected_modout1):
modin = newmodel()
modin_fpath = 'schemaorg/catcherintherye-ugly.md'
literate.parse(open(os.path.join(testresourcepath, modin_fpath)).read(), modin)
FINGERPRINT_RULES = {
SCH_NS('Book'): (
materialize(BF_NS('Instance'),
fprint=[
(BF_NS('isbn'), follow(SCH_NS('isbn'))),
],
)
)
}
TRANSFORM_RULES = {
SCH_NS('name'): link(rel=BF_NS('name')),
SCH_NS('author'): materialize(BF_NS('Person'),
BF_NS('creator'),
vars={
'birthDate': follow(SCH_NS('authorBirthDate'),
origin=var('input-resource'))
},
fprint=[
(BF_NS('name'), target()),
(BF_NS('birthDate'), var('birthDate')),
],
links=[
(BF_NS('name'), target()),
(BF_NS('birthDate'), var('birthDate')),
]
),
}
ppl = generic_pipeline(FINGERPRINT_RULES, TRANSFORM_RULES, LABELIZE_RULES)
modout = ppl.run(input_model=modin)
# Use -s to see this
print('='*10, 'test_basics_1', '='*10)
literate.write(modout)
assert len(modout) == 8
assert len(list(util.all_origins(modout, only_types={BF_NS('Instance')}))) == 1
assert len(list(util.all_origins(modout, only_types={BF_NS('Person')}))) == 1
assert len(list(modout.match(None, BF_NS('birthDate'), '1919-01-01'))) == 1
def expected_modout1():
modout = newmodel()
#literate.parse('''
#''', modout)
return modout
def test_basics_2(testresourcepath):
modin = newmodel()
modin_fpath = 'schemaorg/catcherintherye-ugly.md'
literate.parse(open(os.path.join(testresourcepath, modin_fpath)).read(), modin)
FINGERPRINT_RULES = {
SCH_NS('Book'): (
materialize(var('itype'),
fprint=[
(BF_NS('isbn'), follow(SCH_NS('isbn'))),
],
links=[
(BF_NS('instantiates'),
materialize(BF_NS('Work'),
fprint=[
(BF_NS('name'), follow(SCH_NS('title'))),
(BF_NS('creator'), follow(SCH_NS('author'))),
(BF_NS('language'), var('lang')),
],
links=[('http://instantiated-by', var('@stem'))],
attach=False # Can remove when we have smart sessions to avoid duplicate instantiates links
),
)
],
# Not really necessary; just testing vars in this scenario
vars={
'lang': follow(SCH_NS('inLanguage')),
'itype': BF_NS('Instance')
}
)
)
}
TRANSFORM_RULES = {
# Rule for output resource type of Work or Instance
(SCH_NS('name'), WT, IT): link(rel=BF_NS('name')),
# Rule only for output resource type of Work
(SCH_NS('author'), WT): materialize(BF_NS('Person'),
BF_NS('creator'),
vars={
'birthDate': follow(SCH_NS('authorBirthDate'),
origin=var('input-resource'))
},
fprint=[
# Supplementary type
(VTYPE_REL, SCH_NS('Novelist')),
(BF_NS('name'), target()),
(BF_NS('birthDate'), var('birthDate')),
],
links=[
# Supplementary type
(VTYPE_REL, SCH_NS('Novelist')),
(BF_NS('name'), target()),
(BF_NS('birthDate'), var('birthDate')),
],
preserve_fprint=True,
),
}
ppl = generic_pipeline(FINGERPRINT_RULES, TRANSFORM_RULES, LABELIZE_RULES)
modout = ppl.run(input_model=modin)
# Use -s to see this
print('='*10, 'test_basics_2', '='*10)
literate.write(modout)
#import pprint; pprint.pprint(list(iter(modout)))
assert len(modout) == 15
assert len(list(util.all_origins(modout, only_types={BF_NS('Instance')}))) == 1
assert len(list(util.all_origins(modout, only_types={BF_NS('Work')}))) == 1
assert len(list(util.all_origins(modout, only_types={BF_NS('Person')}))) == 1
assert len(list(modout.match(None, BF_NS('birthDate'), '1919-01-01'))) == 1
def test_basics_4(testresourcepath):
'''
Convert from schema.org to [MusicBrainz scheme](https://musicbrainz.org/doc/MusicBrainz_Database/Schema)
'''
import sys # Uncomment to debug
MB_NS = I('https://musicbrainz.org/doc/MusicBrainz_Database/Schema/')
R_TYP = MB_NS('Release')
RG_TYP = MB_NS('ReleaseGroup')
A_TYP = MB_NS('Artist')
DOC_NS = I('http://example.org/records/')
modin = newmodel()
modin_fpath = 'schemaorg/blackstar.md'
literate.parse(open(os.path.join(testresourcepath, modin_fpath)).read(), modin)
# Hand-add a comment property to the Mos Def resource to test that this value doesn't bleed e.g. to Kweli's output
modin.add(DOC_NS('md'), SCH_NS('comment'), 'test')
FINGERPRINT_RULES = {
SCH_NS('MusicAlbum'): (
materialize(MB_NS('ReleaseGroup'),
fprint=[
(MB_NS('title'), follow(SCH_NS('name'))),
(MB_NS('artist'), follow(SCH_NS('byArtist'), SCH_NS('name'))),
],
links=[
(MB_NS('contains'), materialize(MB_NS('Release'),
fprint=[
(MB_NS('catalogue-number'), var('catnum')),
],
links=[
(MB_NS('catalogue-number'), var('catnum')),
]
))
],
vars={'catnum': follow(SCH_NS('catalogNumber'))},
# debug=sys.stderr, # Uncomment to debug
)
),
SCH_NS('Person'): (
materialize(MB_NS('Artist'),
fprint=[
(MB_NS('name'), var('aname')),
],
links=[
(MB_NS('name'), var('aname')),
(MB_NS('remark'), var('comment')),
],
vars={'aname': follow(SCH_NS('name')), 'comment': follow(SCH_NS('comment'))},
)
)
}
TRANSFORM_RULES = {
(SCH_NS('name'), R_TYP, RG_TYP): link(rel=MB_NS('title')),
(SCH_NS('byArtist'), R_TYP): link(rel=MB_NS('by'), target=lookup('@resource')),
}
# Intentionally shadows the global LABELIZE_RULES
LABELIZE_RULES = {
MB_NS('ReleaseGroup'): follow(MB_NS('title')),
MB_NS('Release'): follow(MB_NS('title')),
MB_NS('Artist'): follow(MB_NS('name'))
}
ppl = generic_pipeline(FINGERPRINT_RULES, TRANSFORM_RULES, LABELIZE_RULES)
modout = ppl.run(input_model=modin)
# Use -s to see this
print('='*10, 'test_basics_4', '='*10)
literate.write(modout)
# import pprint; pprint.pprint(list(iter(modout)))
assert len(modout) == 16
assert len(list(util.all_origins(modout, only_types={MB_NS('ReleaseGroup')}))) == 1
assert len(list(util.all_origins(modout, only_types={MB_NS('ReleaseGroup')}))) == 1
assert len(list(util.all_origins(modout, only_types={MB_NS('Artist')}))) == 2
# assert len(list(modout.match(None, BF_NS('birthDate'), '1919-01-01'))) == 1
# DOC_NS('md') -> I('i5GvPVm7ClA') in the transform
assert [ l[0] for l in modout.match(None, MB_NS('remark'), 'test')] == [I('i5GvPVm7ClA')]
def labelize(self):
'''
Executes a utility rule to create labels in output model for new resources
'''
# XXX Check if there's already a label?
# Apply a common transform strategy using rules defined above
def missed_label(origin, type):
'''
Callback to handle cases where a transform wasn't found to match a link (by relationship) in the input model
'''
warnings.warn(f'No label generated for: {origin}')
labels = self.labelize_helper(LABELIZE_RULES, handle_misses=missed_label)
return True
if __name__ == '__main__':
for rec in INPUT_RECORDS:
ppl = dc_schema_pipeline()
input_model = newmodel()
literate.parse(rec, input_model)
output_model = ppl.run(input_model=input_model)
print('Resulting record Fingerprints:', ppl.fingerprints)
print('Low level JSON dump of output data model: ')
util.jsondump(output_model, sys.stdout)
print('Versa literate form of output: ')
literate.write(output_model, out=sys.stdout)
# from versa.serial import mermaid
# print('Mermaid diagram form of output: ')
# mermaid.write(output_model, out=sys.stdout)