File: base_phrase_segmenter.h

package info (click to toggle)
chromium 139.0.7258.127-1
  • links: PTS, VCS
  • area: main
  • in suites:
  • size: 6,122,068 kB
  • sloc: cpp: 35,100,771; ansic: 7,163,530; javascript: 4,103,002; python: 1,436,920; asm: 946,517; xml: 746,709; pascal: 187,653; perl: 88,691; sh: 88,436; objc: 79,953; sql: 51,488; cs: 44,583; fortran: 24,137; makefile: 22,147; tcl: 15,277; php: 13,980; yacc: 8,984; ruby: 7,485; awk: 3,720; lisp: 3,096; lex: 1,327; ada: 727; jsp: 228; sed: 36
file content (51 lines) | stat: -rw-r--r-- 2,169 bytes parent folder | download | duplicates (8)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
// Copyright 2024 The Chromium Authors
// Use of this source code is governed by a BSD-style license that can be
// found in the LICENSE file.

#ifndef CHROME_RENDERER_ACCESSIBILITY_PHRASE_SEGMENTATION_BASE_PHRASE_SEGMENTER_H_
#define CHROME_RENDERER_ACCESSIBILITY_PHRASE_SEGMENTATION_BASE_PHRASE_SEGMENTER_H_

#include <vector>

#include "chrome/renderer/accessibility/phrase_segmentation/token_boundaries.h"
#include "chrome/renderer/accessibility/phrase_segmentation/tokenized_sentence.h"

// Strategy determines how the phrase segmenter identifies acceptable phrases.
// The phrase segmentation calculation takes one of the strategies below, along
// with a threshold to determine phrase eligibility. For example, if the caller
// wants that phrases can be max 5 words, strategy will be kWords and threshold
// will be 5.
enum class Strategy {
  kWords,
  kCharacters,
};

// An abstract phrase segmenter that breaks an input text into segments
// (phrases) according to a specific algorithm (e.g. smart breaking, random
// breaking, etc.).
class BasePhraseSegmenter {
 public:
  virtual ~BasePhraseSegmenter() = default;
  // Returns a score for every boundary indicating the likelihood of breaking.
  // The inputs are the tokenized sentence, the calculated token boundaries, the
  // strategy for finding the segments ('words' or 'characters') and the
  // threshold for the strategy.
  virtual std::vector<float> CalculatePhraseBoundariesVector(
      const TokenizedSentence& tokenized_sentence,
      const TokenBoundaries& token_boundaries,
      Strategy strategy,
      int threshold) = 0;
};

// Returns the character offsets for splitting the text into phrases.
// The inputs are the tokenized sentence, the calculated token boundaries, the
// strategy for finding the segments ('words' or 'characters') and the
// threshold for the strategy.
std::vector<int> CalculatePhraseBoundaries(
    BasePhraseSegmenter& phrase_segmenter,
    const TokenizedSentence& tokenized_sentence,
    const TokenBoundaries& token_boundaries,
    Strategy strategy,
    int threshold);

#endif  // CHROME_RENDERER_ACCESSIBILITY_PHRASE_SEGMENTATION_BASE_PHRASE_SEGMENTER_H_