Class TextEncoder

tts/utils/text_utils/text_encoder.py:52–118

Base class for converting from ints to/from human readable strings.

class TextEncoder(object):
    """Base class for converting from ints to/from human readable strings."""

    def __init__(self, num_reserved_ids=NUM_RESERVED_TOKENS):
        self._num_reserved_ids = num_reserved_ids

    @property
    def num_reserved_ids(self):
        return self._num_reserved_ids

    def encode(self, s):
        """Transform a human-readable string into a sequence of int ids.

        The ids should be in the range [num_reserved_ids, vocab_size). Ids [0,
        num_reserved_ids) are reserved.

        EOS is not appended.

        Args:
        s: human-readable string to be converted.

        Returns:
        ids: list of integers
        """
        return [int(w) + self._num_reserved_ids for w in s.split()]

    def decode(self, ids, strip_extraneous=False):
        """Transform a sequence of int ids into a human-readable string.

        EOS is not expected in ids.

        Args:
        ids: list of integers to be converted.
        strip_extraneous: bool, whether to strip off extraneous tokens
            (EOS and PAD).

        Returns:
        s: human-readable string.
        """
        if strip_extraneous:
            ids = strip_ids(ids, list(range(self._num_reserved_ids or 0)))
        return " ".join(self.decode_list(ids))

    def decode_list(self, ids):
        """Transform a sequence of int ids into a their string versions.

        This method supports transforming individual input/output ids to their
        string versions so that sequence to/from text conversions can be visualized
        in a human readable format.

        Args:
        ids: list of integers to be converted.

        Returns:
        strs: list of human-readable string.
        """
        decoded_ids = []
        for id_ in ids: