nltk.tokenize.util module

class nltk.tokenize.util.CJKChars

Bases: object

An object that enumerates the code points of the CJK characters as listed on https://en.wikipedia.org/wiki/Basic_Multilingual_Plane#Basic_Multilingual_Plane

This is a Python port of the CJK code point enumerations of Moses tokenizer: https://github.com/moses-smt/mosesdecoder/blob/master/scripts/tokenizer/detokenizer.perl#L309

CJK_Compatibility_Forms = (65072, 65103)

CJK_Compatibility_Ideographs = (63744, 64255)

CJK_Radicals = (11904, 42191)

Hangul_Jamo = (4352, 4607)

Hangul_Syllables = (44032, 55215)

Katakana_Hangul_Halfwidth = (65381, 65500)

Phags_Pa = (43072, 43135)

Supplementary_Ideographic_Plane = (131072, 196607)

ranges = [(4352, 4607), (11904, 42191), (43072, 43135), (44032, 55215), (63744, 64255), (65072, 65103), (65381, 65500), (131072, 196607)]

nltk.tokenize.util.align_tokens(tokens, sentence)

This module attempt to find the offsets of the tokens in s, as a sequence of (start, end) tuples, given the tokens and also the source string.

>>> from nltk.tokenize import TreebankWordTokenizer
>>> from nltk.tokenize.util import align_tokens
>>> s = str("The plane, bound for St Petersburg, crashed in Egypt's "
... "Sinai desert just 23 minutes after take-off from Sharm el-Sheikh "
... "on Saturday.")
>>> tokens = TreebankWordTokenizer().tokenize(s)
>>> expected = [(0, 3), (4, 9), (9, 10), (11, 16), (17, 20), (21, 23),
... (24, 34), (34, 35), (36, 43), (44, 46), (47, 52), (52, 54),
... (55, 60), (61, 67), (68, 72), (73, 75), (76, 83), (84, 89),
... (90, 98), (99, 103), (104, 109), (110, 119), (120, 122),
... (123, 131), (131, 132)]
>>> output = list(align_tokens(tokens, s))
>>> len(tokens) == len(expected) == len(output)  # Check that length of tokens and tuples are the same.
True
>>> expected == list(align_tokens(tokens, s))  # Check that the output is as expected.
True
>>> tokens == [s[start:end] for start, end in output]  # Check that the slices of the string corresponds to the tokens.
True

Parameters

• tokens (list(str)) – The list of strings that are the result of tokenization

• sentence (str) – The original string

Return type

list(tuple(int,int))

nltk.tokenize.util.is_cjk(character)

Python port of Moses’ code to check for CJK character.

>>> CJKChars().ranges
[(4352, 4607), (11904, 42191), (43072, 43135), (44032, 55215), (63744, 64255), (65072, 65103), (65381, 65500), (131072, 196607)]
>>> is_cjk(u'㏾')
True
>>> is_cjk(u'﹟')
False

Parameters

character (char) – The character that needs to be checked.

Returns

bool

nltk.tokenize.util.regexp_span_tokenize(s, regexp)

Return the offsets of the tokens in s, as a sequence of (start, end) tuples, by splitting the string at each successive match of regexp.

>>> from nltk.tokenize.util import regexp_span_tokenize
>>> s = '''Good muffins cost $3.88\nin New York.  Please buy me
... two of them.\n\nThanks.'''
>>> list(regexp_span_tokenize(s, r'\s')) 
[(0, 4), (5, 12), (13, 17), (18, 23), (24, 26), (27, 30), (31, 36),
(38, 44), (45, 48), (49, 51), (52, 55), (56, 58), (59, 64), (66, 73)]

Parameters

• s (str) – the string to be tokenized

• regexp (str) – regular expression that matches token separators (must not be empty)

Return type

iter(tuple(int, int))

nltk.tokenize.util.spans_to_relative(spans)

Return a sequence of relative spans, given a sequence of spans.

>>> from nltk.tokenize import WhitespaceTokenizer
>>> from nltk.tokenize.util import spans_to_relative
>>> s = '''Good muffins cost $3.88\nin New York.  Please buy me
... two of them.\n\nThanks.'''
>>> list(spans_to_relative(WhitespaceTokenizer().span_tokenize(s))) 
[(0, 4), (1, 7), (1, 4), (1, 5), (1, 2), (1, 3), (1, 5), (2, 6),
(1, 3), (1, 2), (1, 3), (1, 2), (1, 5), (2, 7)]

Parameters

spans (iter(tuple(int, int))) – a sequence of (start, end) offsets of the tokens

Return type

iter(tuple(int, int))

nltk.tokenize.util.string_span_tokenize(s, sep)

Return the offsets of the tokens in s, as a sequence of (start, end) tuples, by splitting the string at each occurrence of sep.

>>> from nltk.tokenize.util import string_span_tokenize
>>> s = '''Good muffins cost $3.88\nin New York.  Please buy me
... two of them.\n\nThanks.'''
>>> list(string_span_tokenize(s, " ")) 
[(0, 4), (5, 12), (13, 17), (18, 26), (27, 30), (31, 36), (37, 37),
(38, 44), (45, 48), (49, 55), (56, 58), (59, 73)]

Parameters

• s (str) – the string to be tokenized

• sep (str) – the token separator

Return type

iter(tuple(int, int))

nltk.tokenize.util.xml_escape(text)

This function transforms the input text into an “escaped” version suitable for well-formed XML formatting.

Note that the default xml.sax.saxutils.escape() function don’t escape some characters that Moses does so we have to manually add them to the entities dictionary.

>>> input_str = ''')| & < > ' " ] ['''
>>> expected_output =  ''')| &amp; &lt; &gt; ' " ] ['''
>>> escape(input_str) == expected_output
True
>>> xml_escape(input_str)
')&#124; &amp; &lt; &gt; &apos; &quot; &#93; &#91;'

Parameters

text (str) – The text that needs to be escaped.

Return type

str

nltk.tokenize.util.xml_unescape(text)

This function transforms the “escaped” version suitable for well-formed XML formatting into humanly-readable string.

Note that the default xml.sax.saxutils.unescape() function don’t unescape some characters that Moses does so we have to manually add them to the entities dictionary.

>>> from xml.sax.saxutils import unescape
>>> s = ')| & < > ' " ] ['
>>> expected = ''')| & < > ' " ] ['''
>>> xml_unescape(s) == expected
True

Parameters

text (str) – The text that needs to be unescaped.

Return type

str