diff --git a/spacy/tests/spans/test_merge.py b/spacy/tests/spans/test_merge.py index 2360a0839..406809626 100644 --- a/spacy/tests/spans/test_merge.py +++ b/spacy/tests/spans/test_merge.py @@ -1,7 +1,6 @@ from __future__ import unicode_literals import pytest -@pytest.mark.models def test_merge_tokens(EN): tokens = EN(u'Los Angeles start.') assert len(tokens) == 4 @@ -13,7 +12,6 @@ def test_merge_tokens(EN): assert tokens[0].head.orth_ == 'start' -@pytest.mark.models def test_merge_heads(EN): tokens = EN(u'I found a pilates class near work.') assert len(tokens) == 8 @@ -32,7 +30,6 @@ def test_issue_54(EN): text = u'Talks given by women had a slightly higher number of questions asked (3.2$\pm$0.2) than talks given by men (2.6$\pm$0.1).' tokens = EN(text) -@pytest.mark.models def test_np_merges(EN): text = u'displaCy is a parse tool built with Javascript' tokens = EN(text) @@ -47,3 +44,31 @@ def test_np_merges(EN): merged = tokens.merge(start, end, label, lemma, label) assert merged != None, (start, end, label, lemma) +def test_entity_merge(EN): + tokens = EN(u'Stewart Lee is a stand up comedian who lives in England and loves Joe Pasquale.\n') + assert(len(tokens) == 17) + for ent in tokens.ents: + label, lemma, type_ = (ent.root.tag_, ent.root.lemma_, max(w.ent_type_ for w in ent)) + ent.merge(label, lemma, type_) + # check looping is ok + assert(len(tokens) == 15) + +def test_sentence_update_after_merge(EN): + tokens = EN(u'Stewart Lee is a stand up comedian. He lives in England and loves Joe Pasquale.') + sent1, sent2 = list(tokens.sents) + init_len = len(sent1) + init_len2 = len(sent2) + merge_me = tokens[0:2] + merge_me.merge(u'none', u'none', u'none') + merge_me2 = tokens[-2:] + merge_me2.merge(u'none', u'none', u'none') + assert(len(sent1) == init_len - 1) + assert(len(sent2) == init_len2 - 1) + +def test_subtree_size_check(EN): + tokens = EN(u'Stewart Lee is a stand up comedian who lives in England and loves Joe Pasquale') + sent1 = list(tokens.sents)[0] + init_len = len(list(sent1.root.subtree)) + merge_me = tokens[0:2] + merge_me.merge(u'none', u'none', u'none') + assert(len(list(sent1.root.subtree)) == init_len - 1)