utils/tokenizer.fbs

//
// Copyright (C) 2018 The Android Open Source Project
//
// Licensed under the Apache License, Version 2.0 (the "License");
// you may not use this file except in compliance with the License.
// You may obtain a copy of the License at
//
//      http://www.apache.org/licenses/LICENSE-2.0
//
// Unless required by applicable law or agreed to in writing, software
// distributed under the License is distributed on an "AS IS" BASIS,
// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
// See the License for the specific language governing permissions and
// limitations under the License.
//

// Controls the type of tokenization the model will use for the input text.
namespace libtextclassifier3;
enum TokenizationType : int {
  INVALID_TOKENIZATION_TYPE = 0,

  // Use the internal tokenizer for tokenization.
  INTERNAL_TOKENIZER = 1,

  // Use ICU for tokenization.
  ICU = 2,

  // First apply ICU tokenization. Then identify stretches of tokens
  // consisting only of codepoints in internal_tokenizer_codepoint_ranges
  // and re-tokenize them using the internal tokenizer.
  MIXED = 3,
}

// Role of the codepoints in the range.
namespace libtextclassifier3.TokenizationCodepointRange_;
enum Role : int {
  // Concatenates the codepoint to the current run of codepoints.
  DEFAULT_ROLE = 0,

  // Splits a run of codepoints before the current codepoint.
  SPLIT_BEFORE = 1,

  // Splits a run of codepoints after the current codepoint.
  SPLIT_AFTER = 2,

  // Each codepoint will be a separate token. Good e.g. for Chinese
  // characters.
  TOKEN_SEPARATOR = 3,

  // Discards the codepoint.
  DISCARD_CODEPOINT = 4,

  // Common values:
  // Splits on the characters and discards them. Good e.g. for the space
  // character.
  WHITESPACE_SEPARATOR = 7,
}

// Represents a codepoint range [start, end) with its role for tokenization.
namespace libtextclassifier3;
table TokenizationCodepointRange {
  start:int;
  end:int;
  role:TokenizationCodepointRange_.Role;

  // Integer identifier of the script this range denotes. Negative values are
  // reserved for Tokenizer's internal use.
  script_id:int;
}