def test_sl_pretokenized_conllu():
classla.download('sl', dir=TEST_MODELS_DIR)
nlp = classla.Pipeline('sl', tokenize_pretokenized='conllu', dir=TEST_MODELS_DIR)
conllu_pretokenized = """
# newpar id = 1
# sent_id = 1.1
# text = France Prešeren je rojen v Vrbi.
1 France France _ _ _ _ _ _ _
2 Prešeren Prešeren _ _ _ _ _ _ _
3 je biti _ _ _ _ _ _ _
4 rojen rojen _ _ _ _ _ _ _
5 v v _ _ _ _ _ _ _
6 Vrbi Vrba _ _ _ _ _ _ SpaceAfter=No
7 . . _ _ _ _ _ _ _
"""
doc = nlp(conllu_pretokenized)
assert doc.to_conll().strip() == SL_STANDARD_CONLL
if args.output is None:
output_file_path = args.text_file + '.out'
else:
output_file_path = args.output
# map language code to treebank shorthand
if args.treebank is not None:
treebank_shorthand = args.treebank
else:
treebank_shorthand = default_treebanks[args.language]
# check for models
print('checking for models...')
lang_models_dir = '%s/%s_models' % (args.models_dir, treebank_shorthand)
if not os.path.exists(lang_models_dir):
print('could not find: ' + lang_models_dir)
download(treebank_shorthand,
resource_dir=args.models_dir,
force=args.force_download)
# set up pipeline
pipeline_config = \
dict([(k, v) for k, v in vars(args).items() if k in PROCESSOR_SETTINGS_LIST and v is not None])
pipeline = Pipeline(processors=args.processors,
treebank=treebank_shorthand,
models_dir=args.models_dir,
**pipeline_config)
# build document
print('running pipeline...')
doc = pipeline(open(args.text_file).read())
# write conll to file
doc.write_conll_to_file(output_file_path)
print('done.')
print('results written to: ' + output_file_path)
sentence_df = sentence_df[[
'docId', 'sentenceId', 'tokenId', 'text', 'lemma', 'calcLemma', 'upos',
'xpos', 'ner', 'clID'
]] # leaving out 'misc' for now
return sentence_df, warnings
if __name__ == '__main__':
datasets_files = json.load(open('./data/results/dataset_pairs.json'))
languages = set([
lang for dataset in datasets_files
for lang in datasets_files[dataset].keys()
])
print(languages)
processors = 'tokenize,pos,lemma'
if DOWNLOAD_RESOURCES: # do it once on a new system
for lang in languages:
lang = lang if lang != 'ua' else 'uk'
print(f'Downloading {lang}...')
stanza.download(lang, processors=processors)
classla.download('sl')
classla.download('bg')
tokenizers = {
lang: stanza.Pipeline(lang=lang if lang != 'ua' else 'uk',
processors=processors)
for lang in languages
}
tokenizers['sl'] = classla.Pipeline('sl', processors=processors)
tokenizers['bg'] = classla.Pipeline('bg', processors=processors)
split_documents(datasets_files, tokenizers)
example_sentences = {
"sl": "France Prešeren je rojen v Vrbi.",
"hr": "Ante Starčević rođen je u Velikom Žitniku.",
"sr": "Slobodan Jovanović rođen je u Novom Sadu.",
"bg": "Алеко Константинов е роден в Свищов."
}
if args.lang not in example_sentences:
print(
f'Sorry, but we don\'t have a demo sentence for "{args.lang}" for the moment. Try one of these languages: {list(example_sentences.keys())}'
)
sys.exit(1)
# download the models
classla.download(args.lang, args.models_dir, confirm_if_exists=True)
# set up a pipeline
print('---')
print('Building pipeline...')
pipeline = classla.Pipeline(models_dir=args.models_dir,
lang=args.lang,
use_gpu=(not args.cpu))
# process the document
doc = pipeline(example_sentences[args.lang])
# access nlp annotations
print('')
print('Input: {}'.format(example_sentences[args.lang]))
print("The tokenizer split the input into {} sentences.".format(
len(doc.sentences)))
print('---')
print('tokens of first sentence: ')
example_sentences = {
"sl": "France Prešeren je rojen v Vrbi.",
"hr": "Ante Starčević rođen je u Velikom Žitniku.",
"sr": "Slobodan Jovanović rođen je u Novom Sadu.",
"bg": "Алеко Константинов е роден в Свищов."
}
if args.lang not in example_sentences:
print(
f'Sorry, but we don\'t have a demo sentence for "{args.lang}" for the moment. Try one of these languages: {list(example_sentences.keys())}'
)
sys.exit(1)
# download the models
classla.download(args.lang, args.models_dir)
# set up a pipeline
print('---')
print('Building pipeline...')
pipeline = classla.Pipeline(dir=args.models_dir,
lang=args.lang,
use_gpu=(not args.cpu))
# process the document
doc = pipeline(example_sentences[args.lang])
# access nlp annotations
print('')
print('Input: {}'.format(example_sentences[args.lang]))
print("The tokenizer split the input into {} sentences.".format(
len(doc.sentences)))
print('---')
print('tokens of first sentence: ')
def test_sl_inflectional():
classla.download('sl', dir=TEST_MODELS_DIR)
nlp = classla.Pipeline('sl', pos_use_lexicon=True, dir=TEST_MODELS_DIR)
doc = nlp(SL_STANDARD)
assert doc.to_conll().strip() == SL_STANDARD_CONLL
def test_sl_standard_jos():
classla.download('sl', type='standard_jos', dir=TEST_MODELS_DIR)
nlp = classla.Pipeline('sl', type='standard_jos', dir=TEST_MODELS_DIR)
doc = nlp(SL_STANDARD_JOS)
assert doc.to_conll().strip() == SL_STANDARD_JOS_CONLL
def test_mk_standard():
classla.download('mk', dir=TEST_MODELS_DIR)
nlp = classla.Pipeline('mk', dir=TEST_MODELS_DIR)
doc = nlp(MK_STANDARD)
assert doc.to_conll().strip() == MK_STANDARD_CONLL
def test_sr_nonstandard():
classla.download('sr', type='nonstandard', dir=TEST_MODELS_DIR)
nlp = classla.Pipeline('sr', type='nonstandard', dir=TEST_MODELS_DIR)
doc = nlp(SR_NONSTANDARD)
assert doc.to_conll().strip() == SR_NONSTANDARD_CONLL
def test_all_downloads():
classla.download('sl', dir=TEST_MODELS_DIR)
classla.download('sl', type='standard_jos', dir=TEST_MODELS_DIR)
classla.download('sl', type='nonstandard', dir=TEST_MODELS_DIR)
classla.download('hr', dir=TEST_MODELS_DIR)
classla.download('hr', type='nonstandard', dir=TEST_MODELS_DIR)
classla.download('sr', dir=TEST_MODELS_DIR)
classla.download('sr', type='nonstandard', dir=TEST_MODELS_DIR)
classla.download('bg', dir=TEST_MODELS_DIR)
classla.download('mk', dir=TEST_MODELS_DIR)
