def test_reporter_tokenizer(self):
"""Do we tokenize correctly?"""
self.assertEqual(tokenize('See Roe v. Wade, 410 U. S. 113 (1973)'),
['See', 'Roe', 'v.', 'Wade,', '410', 'U. S.', '113',
'(1973)'])
self.assertEqual(tokenize('Foo bar eats grue, 232 Vet. App. (2003)'),
['Foo', 'bar', 'eats', 'grue,', '232', 'Vet. App.',
'(2003)'])
def test_reporter_tokenizer(self):
"""Do we tokenize correctly?"""
self.assertEqual(
tokenize('See Roe v. Wade, 410 U. S. 113 (1973)'),
['See', 'Roe', 'v.', 'Wade,', '410', 'U. S.', '113', '(1973)'])
self.assertEqual(
tokenize('Foo bar eats grue, 232 Vet. App. (2003)'),
['Foo', 'bar', 'eats', 'grue,', '232', 'Vet. App.', '(2003)'])
# Tests that the tokenizer handles whitespace well. In the past, the
# capital letter P in 5243-P matched the abbreviation for the Pacific
# reporter ("P"), and the tokenizing would be wrong.
self.assertEqual(
tokenize('Failed to recognize 1993 Ct. Sup. 5243-P'),
['Failed', 'to', 'recognize', '1993', 'Ct. Sup.', '5243-P'])
def test_reporter_tokenizer(self):
"""Do we tokenize correctly?"""
self.assertEqual(tokenize('See Roe v. Wade, 410 U. S. 113 (1973)'),
['See', 'Roe', 'v.', 'Wade,', '410', 'U. S.', '113',
'(1973)'])
self.assertEqual(tokenize('Foo bar eats grue, 232 Vet. App. (2003)'),
['Foo', 'bar', 'eats', 'grue,', '232', 'Vet. App.',
'(2003)'])
# Tests that the tokenizer handles whitespace well. In the past, the
# capital letter P in 5243-P matched the abbreviation for the Pacific
# reporter ("P"), and the tokenizing would be wrong.
self.assertEqual(tokenize('Failed to recognize 1993 Ct. Sup. 5243-P'),
['Failed', 'to', 'recognize', '1993', 'Ct. Sup.',
'5243-P'])
def get_citations(text,
html=True,
do_post_citation=True,
do_defendant=True,
disambiguate=True):
if html:
text = get_visible_text(text)
words = reporter_tokenizer.tokenize(text)
citations = []
# Exclude first and last tokens when looking for reporters, because valid
# citations must have a volume before and a page after the reporter.
for i in xrange(0, len(words) - 1):
# Find reporter
if words[i] in (EDITIONS.keys() + VARIATIONS_ONLY.keys()):
citation = extract_base_citation(words, i)
if citation is None:
# Not a valid citation; continue looking
continue
if do_post_citation:
add_post_citation(citation, words)
if do_defendant:
add_defendant(citation, words)
citations.append(citation)
if disambiguate:
# Disambiguate or drop all the reporters
citations = disambiguate_reporters(citations)
for citation in citations:
if not citation.court and is_scotus_reporter(citation):
citation.court = 'scotus'
return citations
def get_citations(text, html=True, do_post_citation=True, do_defendant=True):
if html:
text = get_visible_text(text)
words = reporter_tokenizer.tokenize(text)
citations = []
# Exclude first and last tokens when looking for reporters, because valid
# citations must have a volume before and a page number after the reporter.
for i in xrange(1, len(words) - 1):
# Find reporter
if words[i] in (EDITIONS.keys() + VARIATIONS_ONLY.keys()):
citation = extract_base_citation(words, i)
if citation is None:
# Not a valid citation; continue looking
continue
if do_post_citation:
add_post_citation(citation, words)
if do_defendant:
add_defendant(citation, words)
citations.append(citation)
# Disambiguate or drop all the reporters
citations = disambiguate_reporters(citations)
for citation in citations:
if not citation.court and is_scotus_reporter(citation):
citation.court = 'scotus'
return citations
def test_reporter_tokenizer(self):
"""Do we tokenize correctly?"""
self.assertEqual(
tokenize("See Roe v. Wade, 410 U. S. 113 (1973)"),
["See", "Roe", "v.", "Wade,", "410", "U. S.", "113", "(1973)"],
)
self.assertEqual(
tokenize("Foo bar eats grue, 232 Vet. App. (2003)"),
["Foo", "bar", "eats", "grue,", "232", "Vet. App.", "(2003)"],
)
# Tests that the tokenizer handles whitespace well. In the past, the
# capital letter P in 5243-P matched the abbreviation for the Pacific
# reporter ("P"), and the tokenizing would be wrong.
self.assertEqual(
tokenize("Failed to recognize 1993 Ct. Sup. 5243-P"),
["Failed", "to", "recognize", "1993", "Ct. Sup.", "5243-P"],
)
def get_citations(
text,
html=True,
do_post_citation=True,
do_defendant=True,
disambiguate=True,
):
if html:
text = get_visible_text(text)
words = reporter_tokenizer.tokenize(text)
citations = []
for i in xrange(0, len(words) - 1):
citation_token = words[i]
# CASE 1: Citation token is a reporter (e.g., "U. S.").
# In this case, first try extracting it as a standard, full citation,
# and if that fails try extracting it as a short form citation.
if citation_token in (EDITIONS.keys() + VARIATIONS_ONLY.keys()):
citation = extract_full_citation(words, i)
if citation:
# CASE 1A: Standard citation found, try to add additional data
if do_post_citation:
add_post_citation(citation, words)
if do_defendant:
add_defendant(citation, words)
else:
# CASE 1B: Standard citation not found, so see if this
# reference to a reporter is a short form citation instead
citation = extract_shortform_citation(words, i)
if not citation:
# Neither a full nor short form citation
continue
# CASE 2: Citation token is an "Id." or "Ibid." reference.
# In this case, the citation is simply to the immediately previous
# document, but for safety we won't make that resolution until the
# previous citation has been successfully matched to an opinion.
elif citation_token.lower() in {"id.", "id.,", "ibid."}:
citation = extract_id_citation(words, i)
# CASE 3: Citation token is a "supra" reference.
# In this case, we're not sure yet what the citation's antecedent is.
# It could be any of the previous citations above. Thus, like an Id.
# citation, we won't be able to resolve this reference until the
# previous citations are actually matched to opinions.
elif strip_punct(citation_token.lower()) == "supra":
citation = extract_supra_citation(words, i)
# CASE 4: Citation token is a section marker.
# In this case, it's likely that this is a reference to a non-
# opinion document. So we record this marker in order to keep
# an accurate list of the possible antecedents for id citations.
elif u"§" in citation_token:
citation = NonopinionCitation(match_token=citation_token)
# CASE 5: The token is not a citation.
else:
continue
citations.append(citation)
# Disambiguate each citation's reporter
if disambiguate:
citations = disambiguate_reporters(citations)
citations = remove_address_citations(citations)
# Set each citation's court property to "scotus" by default
for citation in citations:
if (isinstance(citation, Citation) and not citation.court
and is_scotus_reporter(citation)):
citation.court = "scotus"
# Returns a list of citations ordered in the sequence that they appear in
# the document. The ordering of this list is important because we will
# later rely on that order to reconstruct the references of the
# ShortformCitation, SupraCitation, and IdCitation objects.
return citations