def test_fix_rs3filewriter_newlines_in_edus():
"""An RST tree that contains newlines in its EDUs can be converted into
an rs3 file just like a tree without newlines."""
t_good = t('Temporal',
[('N', ['Szeryng subsequently focused on teaching']),
('S', ['before resuming his concert career in 1954.'])])
t_bad = t('Temporal',
[('N', ['Szeryng\nsubsequently\nfocused on\nteaching']),
('S', ['before resuming\nhis concert\ncareer in 1954.'])])
tempfile = NamedTemporaryFile()
RS3FileWriter(t_good, output_filepath=tempfile.name)
produced_good_output_tree = RSTTree(tempfile.name)
tempfile = NamedTemporaryFile()
RS3FileWriter(t_bad, output_filepath=tempfile.name)
produced_bad_output_tree = RSTTree(tempfile.name)
assert produced_good_output_tree.edu_strings == \
produced_bad_output_tree.edu_strings
assert produced_good_output_tree.tree == \
produced_bad_output_tree.tree
def test_read_stagedp_long(fixtures_input_dir):
input_tree = rstc.read_stagedp(
os.path.join(fixtures_input_dir, 'long.stagedp'))
assert isinstance(input_tree, StageDPRSTTree)
tempfile = NamedTemporaryFile()
RS3FileWriter(input_tree, output_filepath=tempfile.name)
produced_output_tree = RSTTree(tempfile.name)
assert input_tree.tree == produced_output_tree.tree
def test_read_dis2_tree(fixtures_input_dir):
input_tree = rstc.read_distree(
os.path.join(fixtures_input_dir, 'rst-example2.dis'))
assert isinstance(input_tree, DisRSTTree)
tempfile = NamedTemporaryFile()
RS3FileWriter(input_tree, output_filepath=tempfile.name)
produced_output_tree = RSTTree(tempfile.name)
assert input_tree.tree == produced_output_tree.tree
def test_read_hilda2(fixtures_input_dir):
input_tree = rstc.read_hilda(os.path.join(fixtures_input_dir,
'long.hilda'))
assert isinstance(input_tree, HILDARSTTree)
tempfile = NamedTemporaryFile()
RS3FileWriter(input_tree, output_filepath=tempfile.name)
produced_output_tree = RSTTree(tempfile.name)
assert input_tree.tree == produced_output_tree.tree
def test_read_hs2015a(fixtures_input_dir):
input_tree = rstc.read_hs2015tree(
os.path.join(fixtures_input_dir, 'short.hs2015'))
assert isinstance(input_tree, HS2015RSTTree)
tempfile = NamedTemporaryFile()
RS3FileWriter(input_tree, output_filepath=tempfile.name)
produced_output_tree = RSTTree(tempfile.name)
assert input_tree.tree == produced_output_tree.tree
def test_read_stagedp_one_edu(fixtures_input_dir):
"""the converter must not crash if the input only consists of one EDU."""
input_tree = rstc.read_stagedp(
os.path.join(fixtures_input_dir, 'one-edu.stagedp'))
assert isinstance(input_tree, StageDPRSTTree)
tempfile = NamedTemporaryFile()
RS3FileWriter(input_tree, output_filepath=tempfile.name)
produced_output_tree = RSTTree(tempfile.name)
assert input_tree.tree == produced_output_tree.tree
def test_read_codra_tree():
input_tree = rstc.read_codra(os.path.join(IN_DIR, 'long.codra'))
assert isinstance(input_tree, CodraRSTTree)
tempfile = NamedTemporaryFile()
RS3FileWriter(input_tree, output_filepath=tempfile.name)
produced_output_tree = RSTTree(tempfile.name)
# there is a difference in the tree objects but I don't know what it is,
# as their pretty-print representations are identical
assert input_tree.tree.pprint() == produced_output_tree.tree.pprint()
def test_rs3filewriter_onesegmenttree():
"""A DGParentedTree with only one segment is correctly converted into an RS3 file and back."""
input_tree = t("N", ["foo"])
expected_output_tree = example2tree('only-one-segment.rs3')
tempfile = NamedTemporaryFile()
RS3FileWriter(input_tree, output_filepath=tempfile.name)
produced_output_tree = RSTTree(tempfile.name)
assert produced_output_tree.edu_strings == produced_output_tree.tree.leaves(
) == ['foo']
assert input_tree == expected_output_tree.tree == produced_output_tree.tree
def test_rs3filewriter_emptytree():
"""An empty DGParentedTree is converted into an empty RS3 file and back."""
input_tree = t("", [])
expected_output_tree = example2tree("empty.rs3")
tempfile = NamedTemporaryFile()
RS3FileWriter(input_tree, output_filepath=tempfile.name)
produced_output_tree = RSTTree(tempfile.name)
assert produced_output_tree.edu_strings == produced_output_tree.tree.leaves(
) == []
assert input_tree == expected_output_tree.tree == produced_output_tree.tree
def test_rs3filewriter_nucsat():
"""A DGParentedTree with one nuc-sat relation is correctly converted into an RS3 file and back."""
input_tree = t("circumstance", [("S", ["foo"]), ("N", ["bar"])])
expected_output_tree = example2tree("foo-bar-circ-foo-to-bar.rs3")
tempfile = NamedTemporaryFile()
RS3FileWriter(input_tree, output_filepath=tempfile.name)
produced_output_tree = RSTTree(tempfile.name)
assert produced_output_tree.edu_strings == produced_output_tree.tree.leaves(
) == ['foo', 'bar']
assert input_tree == expected_output_tree.tree == produced_output_tree.tree
input_tree = t("circumstance", [("N", ["foo"]), ("S", ["bar"])])
expected_output_tree = example2tree("foo-bar-circ-bar-to-foo.rs3")
tempfile = NamedTemporaryFile()
RS3FileWriter(input_tree, output_filepath=tempfile.name)
produced_output_tree = RSTTree(tempfile.name)
assert produced_output_tree.edu_strings == produced_output_tree.tree.leaves(
) == ['foo', 'bar']
assert input_tree == expected_output_tree.tree == produced_output_tree.tree
def test_rs3filewriter_nested():
"""A DGParentedTree with a multinuc relation nested in a nuc-sat relation
is correctly converted into an RS3 file and back."""
input_tree = t('elaboration', [('N', ['eins']),
('S', [('joint', [('N', ['zwei']),
('N', ['drei'])])])])
expected_output_tree = example2tree(
'eins-zwei-drei-(elab-eins-from-(joint-zwei-and-drei).rs3')
tempfile = NamedTemporaryFile()
RS3FileWriter(input_tree, output_filepath=tempfile.name)
produced_output_tree = RSTTree(tempfile.name)
assert produced_output_tree.edu_strings == produced_output_tree.tree.leaves(
) == ['eins', 'zwei', 'drei']
assert input_tree == expected_output_tree.tree == produced_output_tree.tree
def test_rs3filewriter_onesegmenttree_umlauts():
"""A DGParentedTree with only one segment with umlauts is correctly
converted into an RS3 file and back.
"""
edu_string = "Über sein östliches Äußeres"
input_tree = t("N", [edu_string])
expected_output_tree = example2tree('only-one-segment-with-umlauts.rs3')
tempfile = NamedTemporaryFile()
RS3FileWriter(input_tree, output_filepath=tempfile.name)
produced_output_tree = RSTTree(tempfile.name)
assert expected_output_tree.edu_strings == \
produced_output_tree.edu_strings == \
produced_output_tree.tree.leaves() == [edu_string]
assert input_tree == expected_output_tree.tree == produced_output_tree.tree
def test_rs3filewriter_pcc_10575():
"""PCC rs3 file 10575 can be converted rs3 -> dgtree -> rs3' -> dgtree',
without information loss between dgtree and dgtree'.
"""
input_tree = t('interpretation', [('N', [('circumstance', [
('S', ['eins']),
('N', [('contrast', [('N', ['zwei']),
('N', [('cause', [('N', ['drei']),
('S', ['vier'])])])])])
])]), ('S', ['fuenf'])])
expected_output_tree = example2tree('maz-10575-excerpt.rs3')
tempfile = NamedTemporaryFile()
RS3FileWriter(input_tree, output_filepath=tempfile.name)
produced_output_tree = RSTTree(tempfile.name)
assert produced_output_tree.edu_strings == produced_output_tree.tree.leaves(
) == ['eins', 'zwei', 'drei', 'vier', 'fuenf']
assert input_tree == expected_output_tree.tree == produced_output_tree.tree
def example2tree(rs3tree_example_filename,
rs3tree_dir=RS3TREE_DIR,
debug=False):
"""Given the filename of an rs3 file and its directory, return an RSTTree instance of it."""
filepath = os.path.join(rs3tree_dir, rs3tree_example_filename)
return RSTTree(filepath, debug=debug)