tokenizer.py

import unicodedata
from .bert import TOKEN_CLS, TOKEN_SEP, TOKEN_UNK


class Tokenizer(object):

    def __init__(self,
                 token_dict,
                 token_cls=TOKEN_CLS,
                 token_sep=TOKEN_SEP,
                 token_unk=TOKEN_UNK,
                 pad_index=0,
                 cased=False):
        """Initialize tokenizer.

        :param token_dict: A dict maps tokens to indices.
        :param token_cls: The token represents classification.
        :param token_sep: The token represents separator.
        :param token_unk: The token represents unknown token.
        :param pad_index: The index to pad.
        :param cased: Whether to keep the case.
        """
        self._token_dict = token_dict
        self._token_cls = token_cls
        self._token_sep = token_sep
        self._token_unk = token_unk
        self._pad_index = pad_index
        self._cased = cased

    @staticmethod
    def _truncate(first_tokens, second_tokens=None, max_len=None):
        if max_len is None:
            return

        if second_tokens is not None:
            while True:
                total_len = len(first_tokens) + len(second_tokens)
                if total_len <= max_len - 3:  # 3 for [CLS] .. tokens_a .. [SEP] .. tokens_b [SEP]
                    break
                if len(first_tokens) > len(second_tokens):
                    first_tokens.pop()
                else:
                    second_tokens.pop()
        else:
            del first_tokens[max_len - 2:]  # 2 for [CLS] .. tokens .. [SEP]

    def _pack(self, first_tokens, second_tokens=None):
        first_packed_tokens = [self._token_cls] + first_tokens + [self._token_sep]
        if second_tokens is not None:
            second_packed_tokens = second_tokens + [self._token_sep]
            return first_packed_tokens + second_packed_tokens, len(first_packed_tokens), len(second_packed_tokens)
        else:
            return first_packed_tokens, len(first_packed_tokens), 0

    def _convert_tokens_to_ids(self, tokens):
        unk_id = self._token_dict.get(self._token_unk)
        return [self._token_dict.get(token, unk_id) for token in tokens]

    def tokenize(self, first, second=None):
        first_tokens = self._tokenize(first)
        second_tokens = self._tokenize(second) if second is not None else None
        tokens, _, _ = self._pack(first_tokens, second_tokens)
        return tokens

    def encode(self, first, second=None, max_len=None):
        first_tokens = self._tokenize(first)
        second_tokens = self._tokenize(second) if second is not None else None
        self._truncate(first_tokens, second_tokens, max_len)
        tokens, first_len, second_len = self._pack(first_tokens, second_tokens)

        token_ids = self._convert_tokens_to_ids(tokens)
        segment_ids = [0] * first_len + [1] * second_len

        if max_len is not None:
            pad_len = max_len - first_len - second_len
            token_ids += [self._pad_index] * pad_len
            segment_ids += [0] * pad_len

        return token_ids, segment_ids

    def _tokenize(self, text):
        if not self._cased:
            text = unicodedata.normalize('NFD', text)
            text = ''.join([ch for ch in text if unicodedata.category(ch) != 'Mn'])
            text = text.lower()
        spaced = ''
        for ch in text:
            if self._is_punctuation(ch) or self._is_cjk_character(ch):
                spaced += ' ' + ch + ' '
            elif self._is_space(ch):
                spaced += ' '
            elif ord(ch) == 0 or ord(ch) == 0xfffd or self._is_control(ch):
                continue
            else:
                spaced += ch
        tokens = []
        for word in spaced.strip().split():
            tokens += self._word_piece_tokenize(word)
        return tokens

    def _word_piece_tokenize(self, word):
        if word in self._token_dict:
            return [word]
        tokens = []
        start, stop = 0, 0
        while start < len(word):
            stop = len(word)
            while stop > start:
                sub = word[start:stop]
                if start > 0:
                    sub = '##' + sub
                if sub in self._token_dict:
                    break
                stop -= 1
            if start == stop:
                stop += 1
            tokens.append(sub)
            start = stop
        return tokens

    @staticmethod
    def _is_punctuation(ch):
        code = ord(ch)
        return 33 <= code <= 47 or \
            58 <= code <= 64 or \
            91 <= code <= 96 or \
            123 <= code <= 126 or \
            unicodedata.category(ch).startswith('P')

    @staticmethod
    def _is_cjk_character(ch):
        code = ord(ch)
        return 0x4E00 <= code <= 0x9FFF or \
            0x3400 <= code <= 0x4DBF or \
            0x20000 <= code <= 0x2A6DF or \
            0x2A700 <= code <= 0x2B73F or \
            0x2B740 <= code <= 0x2B81F or \
            0x2B820 <= code <= 0x2CEAF or \
            0xF900 <= code <= 0xFAFF or \
            0x2F800 <= code <= 0x2FA1F

    @staticmethod
    def _is_space(ch):
        return ch == ' ' or ch == '\n' or ch == '\r' or ch == '\t' or \
            unicodedata.category(ch) == 'Zs'

    @staticmethod
    def _is_control(ch):
        return unicodedata.category(ch).startswith('C')