def testWalk(self):
d = Dictionary([('key', 'the term', 42)], DictionaryTests.tokenizer)
s = "Here is the term we're looking for."
tokens = [s[start:end] for start, end, tag, ortho in DictionaryTests.tokenizer.tokenize(s)]
result = list(d.walk(tokens))
O = Dictionary.O
B = Dictionary.B % 'key'
I = Dictionary.I % 'key'
expected = [O, O, B, I, O, O, O, O, O, O]
self.assertEqual(result, expected)
def testCapitalizationAlts(self):
d = Dictionary(
[('NEUROD1', 'NEUROD', 100),
('NEUROD2', 'NEUROD2', 100)],
DictionaryTests.tokenizer
)
s = "Transfection of vectors expressing neuroD and neuroD2 into P19 cells."
tokens = [s[start:end] for start, end, tag, ortho in DictionaryTests.tokenizer.tokenize(s)]
result = list(d.walk(tokens))
O = Dictionary.O
B = Dictionary.B % 'NEUROD'
I = Dictionary.I % 'NEUROD'
expected = [O, O, O, O, B + "1", I + "1", O, B + "2", I + "2", I + "2", O, O, O, O, O]
self.assertEqual(result, expected)
def testFullWalk(self):
d = Dictionary(
[('alt', 'the term we', 84),
('key', 'the term we', 42),
('apo', "'", 1),
('part', 'term we', 21)],
DictionaryTests.tokenizer
)
s = "Here is the term we're looking for."
tokens = [s[start:end] for start, end, tag, ortho in DictionaryTests.tokenizer.tokenize(s)]
result = list(d.walk(tokens))
O = Dictionary.O
B = Dictionary.B % 'key'
I = Dictionary.I % 'key'
A = Dictionary.B % 'apo'
expected = [O, O, B, I, I, A, O, O, O, O]
self.assertEqual(result, expected)
def testExamples(self):
d = Dictionary(
[('NR1D1', 'rev erb α', 1),
('NR1D1', 'rev erb alpha', 1),
('PPARA', 'PPAR', 1)],
DictionaryTests.tokenizer
)
O = Dictionary.O
B = Dictionary.B
I = Dictionary.I
sentences = [
("A functional Rev-erb alpha responsive element located in the human Rev-erb alpha promoter mediates a repressing activity.",
[O, O, B % 'NR1D1', I % 'NR1D1', I % 'NR1D1', O, O, O, O, O, O, B % 'NR1D1', I % 'NR1D1', I % 'NR1D1', O, O, O, O, O, O]),
("A positive PPAR-response element in the human apoA-I promoter nonfunctional in rats.",
[O, O, B % 'PPARA', O, O, O, O, O, O, O, O, O, O, O, O, O])
]
for s, e in sentences:
r = list(d.walk([s[start:end] for start, end, tag, ortho in DictionaryTests.tokenizer.tokenize(s)]))
self.assertEqual(len(r), len(e))
self.assertEqual(r, e)
def testCreateDictionary(self):
d = Dictionary([('key', 'The Term', 42, 21)], DictionaryTests.tokenizer)
n = Node(The=Node(Term=Node(([42, 21], 'key'))))
self.assertEqual(d.root, n)
else:
parser.error("unknown output option " + args.output)
method = lambda *args: None
try:
pos_tagger = GeniaTagger()
ner_tagger = NerSuite(args.model)
qualifier_list = [l.strip() for l in args.qranks]
raw_dict_data = [
dictionaryReader(d, qualifier_list, args.separator)
for d in args.dictionary
]
# a tokenizer that skips Unicode Categories Zs and Pd:
tokenizer = WordTokenizer(skipTags={'space'}, skipOrthos={'e'})
dictionaries = [
Dictionary(stream, tokenizer) for stream in raw_dict_data
]
logging.info("initialized %s dictionaries", len(dictionaries))
lst = [dictionaries, tokenizer, pos_tagger, ner_tagger]
kwds = dict(sep=args.separator,
tag_all_nouns=args.nouns,
use_greek_letters=args.greek)
if args.files:
lst.append(args.files)
else:
lst.append([sys.stdin])
method(*lst, **kwds)
del ner_tagger
