mindspore.dataset.text.RegexTokenizer

class mindspore.dataset.text.RegexTokenizer(delim_pattern, keep_delim_pattern='', with_offsets=False)[source]

Tokenize a scalar tensor of UTF-8 string by regex expression pattern.

See https://unicode-org.github.io/icu/userguide/strings/regexp.html for supported regex pattern.

Note

RegexTokenizer is not supported on Windows platform yet.

Parameters

  • delim_pattern (str) – The pattern of regex delimiters. The original string will be split by matched elements.

  • keep_delim_pattern (str, optional) – The string matched by 'delim_pattern' can be kept as a token if it can be matched by 'keep_delim_pattern'. The default value is an empty str which means that delimiters will not be kept as an output token. Default: ''.

  • with_offsets (bool, optional) – Whether to output the start and end offsets of each token in the original string. Default: False .

Raises

  • TypeError – If delim_pattern is not of type string.

  • TypeError – If keep_delim_pattern is not of type string.

  • TypeError – If with_offsets is not of type bool.

Supported Platforms:

CPU

Examples

>>> import mindspore.dataset as ds
>>> import mindspore.dataset.text as text
>>>
>>> # Use the transform in dataset pipeline mode
>>> numpy_slices_dataset = ds.NumpySlicesDataset(data=['Welcome  |,  To  |,  BeiJing!'],
...                                              column_names=["text"])
>>>
>>> # 1) If with_offsets=False, default output is one column {["text", dtype=str]}
>>> delim_pattern = r"[ |,]"
>>> tokenizer_op = text.RegexTokenizer(delim_pattern, with_offsets=False)
>>> numpy_slices_dataset = numpy_slices_dataset.map(operations=tokenizer_op)
>>> for item in numpy_slices_dataset.create_dict_iterator(num_epochs=1, output_numpy=True):
...     print(item["text"])
['Welcome' 'To' 'BeiJing!']
>>>
>>> # 2) If with_offsets=True, then output three columns {["token", dtype=str],
>>> #                                                     ["offsets_start", dtype=uint32],
>>> #                                                     ["offsets_limit", dtype=uint32]}
>>> numpy_slices_dataset = ds.NumpySlicesDataset(data=['Welcome  |,  To  |,  BeiJing!'],
...                                              column_names=["text"])
>>> tokenizer_op = text.RegexTokenizer(delim_pattern, with_offsets=True)
>>> numpy_slices_dataset = numpy_slices_dataset.map(
...     operations=tokenizer_op,
...     input_columns=["text"],
...     output_columns=["token", "offsets_start", "offsets_limit"])
>>> for item in numpy_slices_dataset.create_dict_iterator(num_epochs=1, output_numpy=True):
...     print(item["token"], item["offsets_start"], item["offsets_limit"])
['Welcome' 'To' 'BeiJing!'] [ 0 13 21] [ 7 15 29]
>>>
>>> # Use the transform in eager mode
>>> data = 'Welcome     To   BeiJing!'
>>> output = text.RegexTokenizer(delim_pattern="To", keep_delim_pattern="To", with_offsets=True)(data)
>>> print(output)
(array(['Welcome     ', 'To', '   BeiJing!'], dtype='<U12'),
array([ 0, 12, 14], dtype=uint32), array([12, 14, 25], dtype=uint32))
Tutorial Examples: