Mercurial > hgrepos > Python2 > PyMuPDF

comparison mupdf-source/thirdparty/tesseract/src/training/unicharset/validate_grapheme.h @ 3:2c135c81b16c

Find changesets by keywords (author, files, the commit message), revision number or hash, or revset expression.
MERGE: upstream PyMuPDF 1.26.4 with MuPDF 1.26.7
author Franz Glasner <fzglas.hg@dom66.de>
date Mon, 15 Sep 2025 11:44:09 +0200
parents b50eed0cc0ef
children
comparison
equal deleted inserted replaced
0:6015a75abc2d 3:2c135c81b16c
1 #ifndef TESSERACT_TRAINING_VALIDATE_GRAPHEME_H_
2 #define TESSERACT_TRAINING_VALIDATE_GRAPHEME_H_
3
4 #include "validator.h"
5
6 namespace tesseract {
7
8 // Subclass of Validator that validates and segments generic unicode into
9 // grapheme clusters, including Latin with diacritics.
10 class ValidateGrapheme : public Validator {
11 public:
12 ValidateGrapheme(ViramaScript script, bool report_errors) : Validator(script, report_errors) {}
13 ~ValidateGrapheme() override = default;
14
15 protected:
16 // Consumes the next Grapheme in codes_[codes_used_++...] and copies it to
17 // parts_ and output_. Returns true if a valid Grapheme was consumed,
18 // otherwise does not increment codes_used_.
19 bool ConsumeGraphemeIfValid() override;
20 // Returns the CharClass corresponding to the given Unicode ch.
21 CharClass UnicodeToCharClass(char32 ch) const override;
22
23 private:
24 // Helper returns true if the sequence prev_ch,ch is invalid.
25 bool IsBadlyFormed(char32 prev_ch, char32 ch);
26 // Helper returns true if the sequence prev_ch,ch is an invalid Indic vowel.
27 static bool IsBadlyFormedIndicVowel(char32 prev_ch, char32 ch);
28 // Helper returns true if the sequence prev_ch,ch is invalid Thai.
29 static bool IsBadlyFormedThai(char32 prev_ch, char32 ch);
30 };
31
32 } // namespace tesseract
33
34 #endif // TESSERACT_TRAINING_VALIDATE_GRAPHEME_H_