def testSplitWithOffsetsWithDifferentEncodings(self, encoding, texts):
expected_codepoints = _nested_splitchars(texts, encoding)
expected_offsets = _nested_offsets(texts, encoding)
input_tensor = constant_op.constant(_nested_encode(texts, encoding))
result = ragged.unicode_split_with_offsets(input_tensor, encoding)
self.assertRaggedEqual(expected_codepoints, result[0])
self.assertRaggedEqual(expected_offsets, result[1])
def testErrorModesWithOffsets(self,
expected=None,
expected_offsets=None,
**args):
result = ragged.unicode_split_with_offsets(**args)
self.assertRaggedEqual(expected, result[0])
self.assertRaggedEqual(expected_offsets, result[1])
def testVectorSplitWithOffset(self):
text = constant_op.constant([u"仅今年前".encode("UTF-8"), b"hello"])
chars, starts = ragged.unicode_split_with_offsets(text, "UTF-8")
expected_chars = [[c.encode("UTF-8") for c in u"仅今年前"],
[c.encode("UTF-8") for c in u"hello"]]
self.assertRaggedEqual(chars, expected_chars)
self.assertRaggedEqual(starts, [[0, 3, 6, 9], [0, 1, 2, 3, 4]])
def testBasicSplitWithOffsets(self, texts, ragged_rank=None):
input_tensor = ragged.constant_value(_nested_encode(texts, "UTF-8"),
ragged_rank=ragged_rank,
dtype=bytes)
result = ragged.unicode_split_with_offsets(input_tensor, "UTF-8")
expected_codepoints = _nested_splitchars(texts, "UTF-8")
expected_offsets = _nested_offsets(texts, "UTF-8")
self.assertRaggedEqual(expected_codepoints, result[0])
self.assertRaggedEqual(expected_offsets, result[1])
def testDocstringExamples(self):
texts = [s.encode("utf8") for s in [u"G\xf6\xf6dnight", u"\U0001f60a"]]
codepoints1 = ragged.unicode_split(texts, "UTF-8")
codepoints2, offsets = ragged.unicode_split_with_offsets(
texts, "UTF-8")
self.assertRaggedEqual(codepoints1, [[
b"G", b"\xc3\xb6", b"\xc3\xb6", b"d", b"n", b"i", b"g", b"h", b"t"
], [b"\xf0\x9f\x98\x8a"]])
self.assertRaggedEqual(codepoints2, [[
b"G", b"\xc3\xb6", b"\xc3\xb6", b"d", b"n", b"i", b"g", b"h", b"t"
], [b"\xf0\x9f\x98\x8a"]])
self.assertRaggedEqual(offsets, [[0, 1, 3, 5, 6, 7, 8, 9, 10], [0]])
def testScalarSplitWithOffset(self):
text = constant_op.constant(u"仅今年前".encode("UTF-8"))
chars, starts = ragged.unicode_split_with_offsets(text, "UTF-8")
self.assertAllEqual(chars, [c.encode("UTF-8") for c in u"仅今年前"])
self.assertAllEqual(starts, [0, 3, 6, 9])